CN107277500B - The treating method and apparatus that video compares - Google Patents

The treating method and apparatus that video compares Download PDF

Info

Publication number
CN107277500B
CN107277500B CN201710502377.0A CN201710502377A CN107277500B CN 107277500 B CN107277500 B CN 107277500B CN 201710502377 A CN201710502377 A CN 201710502377A CN 107277500 B CN107277500 B CN 107277500B
Authority
CN
China
Prior art keywords
pel
vision signal
frame image
comparison
characteristic value
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201710502377.0A
Other languages
Chinese (zh)
Other versions
CN107277500A (en
Inventor
许钢鸣
祥祖军
黄振川
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
China Central TV Station
Original Assignee
China Central TV Station
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by China Central TV Station filed Critical China Central TV Station
Priority to CN201710502377.0A priority Critical patent/CN107277500B/en
Publication of CN107277500A publication Critical patent/CN107277500A/en
Application granted granted Critical
Publication of CN107277500B publication Critical patent/CN107277500B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N17/00—Diagnosis, testing or measuring for television systems or their details
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N17/00—Diagnosis, testing or measuring for television systems or their details
    • H04N17/004—Diagnosis, testing or measuring for television systems or their details for digital television systems
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N5/00—Details of television systems
    • H04N5/04—Synchronising

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • General Health & Medical Sciences (AREA)
  • Television Systems (AREA)
  • Testing, Inspecting, Measuring Of Stereoscopic Televisions And Televisions (AREA)

Abstract

The present invention provides a kind for the treatment of method and apparatus that video compares, wherein, this method comprises: for first via vision signal to be compared and the second tunnel vision signal, following procedure is executed respectively: for each frame image of vision signal, according to the image feature value of each pixel of current frame image, the fit characteristic value of each comparison pel of current frame image is determined；Find the video frame synchronization point of first via vision signal and the second tunnel vision signal；For each frame image since video frame synchronization point, determine the fit characteristic value of each comparison pel of the first via vision signal on current frame image, each difference between the fit characteristic value for respectively comparing pel of the second tunnel vision signal determines whether first via vision signal and the second tunnel vision signal are consistent on each frame image according to each difference.The video analysis method accurate and effective of offer, can effectively identify whether each frame image of vision signal has occurred exception, effectively identify whether vision signal is abnormal.

Description

The treating method and apparatus that video compares
Technical field
The present invention relates to the treating method and apparatus that broadcasting television technology field more particularly to a kind of video compare.
Background technique
With the fast development of TV tech and video media technology, for the view played in field of broadcast televisions The safety of frequency is increasingly paid attention to, and the grade of the safe broadcast of video is also increasingly stringenter.To need the view to broadcasting Frequency is monitored, and then can be found that the abnormal problem of vision signal within the very short time, to be handled.
It in the prior art, is the conditions such as still frame based on video, Hei Chang, colour bar, color field for the monitoring of video, to video Broadcast state judged judge whether vision signal abnormal, whether the broadcast state to judge video abnormal.
However in the prior art, for the monitoring of video, the vision signal all the way of transmission can be detected, can only be being examined When measuring video and still frame or black field or colour bar occur, determine that vision signal is broadcast shape that is abnormal, and then determining video State is abnormal.Such monitoring mode, is only able to detect still frame, Hei Chang, colour bar, these abnormalities of color field of video, and is directed to It cannot detected in the vision signal unusual condition that other are likely to occur, and then when the exception for analyzing vision signal, it should Method timeliness is lower.
Summary of the invention
The present invention provides the treating method and apparatus that a kind of video compares, can to solve to be directed to other in the prior art Can occur vision signal unusual condition cannot detected, and then for analyze vision signal exception when, existing method and The lower problem of when property.
It is an aspect of the present invention to provide the processing methods that a kind of video compares, comprising:
For first via vision signal to be compared and the second tunnel vision signal, following procedure is executed respectively: for described Each frame image of vision signal determines each of current frame image according to the image feature value of each pixel of current frame image Compare the fit characteristic value of pel；Wherein, each comparison pel is that a frame image is divided into the region of predetermined number and is obtained It arrives；
The fit characteristic value for determining each comparison pel of the first via vision signal on each frame image, with described Whether the fit characteristic value of each comparison pel of the two tunnel vision signals on each frame image is identical, to find the first via view The video frame synchronization point of frequency signal and second tunnel vision signal, wherein the video frame synchronization point characterizes the first via Vision signal and second tunnel vision signal from the video frame synchronization point be initially synchronous；
For each frame image since the video frame synchronization point, determine the first via vision signal in present frame The fit characteristic value of each comparison pel on image, with second tunnel vision signal in the present frame with first via vision signal Each difference between the fit characteristic value of each comparison pel on image corresponding to image, and institute is determined according to each difference It is whether consistent on each frame image to state first via vision signal and second tunnel vision signal.
Another aspect of the present invention is to provide a kind of processing unit that video compares, comprising:
Determining module, for executing respectively following for first via vision signal to be compared and the second tunnel vision signal Process: for each frame image of the vision signal, according to the image feature value of each pixel of current frame image, determination is worked as The fit characteristic value of each comparison pel of prior image frame；Wherein, each comparison pel is that a frame image is divided into default Obtained from several regions；
Analysis module, the fitting for determining each comparison pel of the first via vision signal on each frame image are special Value indicative, it is whether identical as the fit characteristic value that respectively compares pel of second tunnel vision signal on each frame image, to seek Look for the video frame synchronization point of the first via vision signal Yu second tunnel vision signal, wherein the video frame synchronization point Characterize the first via vision signal and second tunnel vision signal from the video frame synchronization point be initially synchronous；
Comparison module, for determining the first via view for each frame image since the video frame synchronization point The fit characteristic value of each comparison pel of the frequency signal on current frame image, regards with second tunnel vision signal with the first via Each difference between the fit characteristic value of each comparison pel on image corresponding to the current frame image of frequency signal, and according to institute It states each difference and determines whether the first via vision signal and second tunnel vision signal are consistent on each frame image.
The solution have the advantages that: by dividing for first via vision signal to be compared and the second tunnel vision signal It Zhi Hang following procedure: for each frame image of vision signal, according to the image feature value of each pixel of current frame image, Determine the fit characteristic value of each comparison pel of current frame image；Wherein, it is default for a frame image to be divided into respectively to compare pel Obtained from the region of number；Determine the fit characteristic value of each comparison pel of the first via vision signal on each frame image, It is whether identical as the fit characteristic value that respectively compares pel of the second tunnel vision signal on each frame image, to find first via view The video frame synchronization point of frequency signal and the second tunnel vision signal, wherein video frame synchronization point characterizes first via vision signal and the Two tunnel vision signals from the video frame synchronization point be initially synchronous；For each frame image since video frame synchronization point, The fit characteristic value for determining each comparison pel of the first via vision signal on current frame image, with the second tunnel vision signal with Each difference between the fit characteristic value of each comparison pel on image corresponding to the current frame image of first via vision signal, And determine whether first via vision signal and the second tunnel vision signal are consistent on each frame image according to each difference.To provide The comparison that two-path video signal carries out vision signal is analyzed, can more accurately be supervised by a kind of new video comparison method It measures and whether extremely to broadcast vision signal；Also, the division in region is carried out for each frame image of vision signal, It obtains at least one and compares pel, the consistency of each frame image is analyzed using the fit characteristic value for comparing pel, is mentioned The video analysis method accurate and effective of confession, can effectively identify whether each frame image of vision signal has occurred exception, And then effectively identify whether vision signal is abnormal.
Detailed description of the invention
Fig. 1 is the flow chart for the processing method that the video that the embodiment of the present invention one provides compares；
Fig. 2 is the flow chart one for the processing method that video provided by Embodiment 2 of the present invention compares；
Fig. 3 is the schematic diagram of the basic pel in the processing method that video provided by Embodiment 2 of the present invention compares；
Fig. 4 is the schematic diagram one of the comparison pel in the processing method that video provided by Embodiment 2 of the present invention compares；
Fig. 5 is the schematic diagram two of the comparison pel in the processing method that video provided by Embodiment 2 of the present invention compares；
Fig. 6 is the process of the searching video frame synchronization point in the processing method that video provided by Embodiment 2 of the present invention compares Figure；
Fig. 7 is the flowchart 2 for the processing method that video provided by Embodiment 2 of the present invention compares；
Fig. 8 is the structural schematic diagram for the processing unit that the video that the embodiment of the present invention three provides compares；
Fig. 9 is the structural schematic diagram for the processing unit that the video that the embodiment of the present invention four provides compares.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is A part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art Every other embodiment obtained without making creative work, shall fall within the protection scope of the present invention.
Fig. 1 is the flow chart for the processing method that the video that the embodiment of the present invention one provides compares, as shown in Figure 1, this implementation The method of example, comprising:
Step 101, for first via vision signal to be compared and the second tunnel vision signal, execute following procedure respectively: Current frame image is determined according to the image feature value of each pixel of current frame image for each frame image of vision signal Each comparison pel fit characteristic value；Wherein, respectively comparing pel is that a frame image is divided into the region of predetermined number and is obtained It arrives.
In the present embodiment, specifically, the executing subject of the present embodiment can be the processing unit or clothes of video comparison Business device or any other can execute the device or system of the method for the present embodiment.
When transmitting the vision signal of TV, transmitting two paths vision signal is needed, is divided into main road vision signal and standby Road vision signal, under normal circumstances, main road vision signal is as vision signal to be played.It in the present embodiment, can be by master Road vision signal is known as first via vision signal, and standby road vision signal is known as the second tunnel vision signal；Alternatively, first via video Signal, the second tunnel vision signal are all main road vision signal or first via vision signal, the second tunnel road vision signal Dou Shibei Vision signal；The comparison analysis for carrying out vision signal to two-path video signal is gone, can more accurately monitor to need to carry out Whether abnormal broadcast vision signal.
For first via vision signal to be compared and the second tunnel vision signal, require to determine respectively respective each A fit characteristic value for comparing pel.
For first via vision signal, each frame image for first via vision signal is needed, each frame image is drawn It is divided into the region of predetermined number, and then has marked off the comparison pel of predetermined number on each frame image, is i.e. each region is One comparison pel；For each frame image, then according to the image feature value of each pixel in current frame image, meter Calculate the fit characteristic value that pel is respectively compared in current frame image, wherein a comparison pel has a fit characteristic value.Its In, for a pixel, the image feature value of a pixel includes Y characteristic value, U characteristic value and V characteristic value；Y Characteristic value, U characteristic value and V characteristic value are according to obtained from colour coding method in the prior art.
Likewise, each frame image for the second tunnel vision signal is needed, by each frame for the second tunnel vision signal Image is divided into the region of predetermined number, and then the comparison pel of predetermined number has been marked off on each frame image, i.e., each Region is a comparison pel；It is then special according to the image of each pixel in current frame image for each frame image Value indicative calculates the fit characteristic value that pel is respectively compared in current frame image, wherein a comparison pel has a fitting special Value indicative.Wherein, for a pixel, the image feature value of a pixel includes Y characteristic value, U characteristic value and V special Value indicative.
Step 102, the fit characteristic value for determining each comparison pel of the first via vision signal on each frame image, with Whether the fit characteristic value of each comparison pel of the two tunnel vision signals on each frame image is identical, to find first via video letter Video frame synchronization point number with the second tunnel vision signal, wherein video frame synchronization point characterizes first via vision signal and the second tunnel Vision signal from the video frame synchronization point be initially synchronous.
In the present embodiment, specifically, first by each frame image of first via vision signal and the second tunnel vision signal Each frame image be compared, specifically, by each comparison pel of the first via vision signal on each frame image Fit characteristic value, be compared with the fit characteristic value that respectively compares pel of the second tunnel vision signal on each frame image point Analysis, goes the fit characteristic value for judging each comparison pel of the first via vision signal on each frame image, believes with the second road video Whether the fit characteristic value of each comparison pel number on each frame image is identical, and then judges the every of first via vision signal Whether one frame image is respectively identical with each frame image of the second tunnel vision signal.It is compared to multiple image After analysis, search out in first via vision signal and the second tunnel vision signal when continuous Q frame image is identical, so that it may Have found video frame synchronization point with determination, the video frame synchronization point characterize first via vision signal and the second tunnel vision signal from this That frame image of video frame synchronization point characterization is initially synchronous.
For example, video frame synchronization point indicates the i-th frame image from first via vision signal, the second tunnel vision signal Jth frame image starts, and first via vision signal is synchronous with the second tunnel vision signal；Wherein, i, j are positive integer.
Step 103, for each frame image since video frame synchronization point, determine first via vision signal in present frame The fit characteristic value of each comparison pel on image, with the second tunnel vision signal in the current frame image with first via vision signal Each difference between the fit characteristic value of each comparison pel on corresponding image, and first via video is determined according to each difference Whether signal and the second tunnel vision signal are consistent on each frame image.
In the present embodiment, it specifically, after determining video frame synchronization point, is directed to from the video frame synchronization point and opens The each frame image to begin is fitted the analysis of characteristic value, go to determine first via vision signal and the second tunnel vision signal from this It is whether consistent on each frame image that video frame synchronization point starts.
For example, video frame synchronization point indicates the i-th frame image from first via vision signal, the jth of the second tunnel vision signal Frame image starts, and first via vision signal is synchronous with the second tunnel vision signal；It so can be from first via vision signal The i-th frame image start to analyze first via vision signal, to second since the jth frame image of the second tunnel vision signal Road vision signal is analyzed, and the i-th frame image of first via vision signal and the jth frame image of the second tunnel vision signal are carried out Analysis is compared, the i+1 frame image of first via vision signal and+1 frame image of jth of the second tunnel vision signal are analyzed, I-th+2 frame image of first via vision signal and+2 frame image of jth of the second tunnel vision signal are analyzed, and so on. Wherein, i, j are positive integer.
Specifically, it for each frame image since video frame synchronization point, calculates first via vision signal and is working as The fit characteristic value of each comparison pel on prior image frame, with the second tunnel vision signal in the present frame with first via vision signal Each difference between the fit characteristic value of each comparison pel on image corresponding to image.For example, video frame synchronization point Indicate that the i-th frame image from first via vision signal, the jth frame image of the second tunnel vision signal start, first via vision signal It is synchronous with the second tunnel vision signal；Pel is compared for the i-th frame image of first via vision signal each, for the The current comparison pel of i-th frame image of vision signal all the way calculates the fit characteristic of the current comparison pel of the i-th frame image Value, and with the fit characteristic for comparing pel corresponding with current comparison chart member on the jth frame image of the second tunnel vision signal Difference between value；Then, for each comparison pel of the i+1 frame image of first via vision signal, for the first via The fit characteristic value of the current comparison pel of the i+1 frame image of vision signal, calculates the current comparison chart of the i+1 frame image Member, and it is special with the fitting for comparing pel corresponding with current comparison chart member on+1 frame image of jth of the second tunnel vision signal Difference between value indicative；And so on.
Then, it for each frame image since video frame synchronization point, is regarded according to first via vision signal and the second tunnel Each difference of the frequency signal on corresponding frame image goes analysis first via vision signal corresponding at this with the second tunnel vision signal A frame image on whether be consistent.
For example, for each comparison pel of the i-th frame image of first via vision signal, for first via video The current comparison pel of i-th frame image of signal calculates the fit characteristic value of the current comparison pel of the i-th frame image, with Between the fit characteristic value for comparing pel corresponding current comparison chart member on the jth frame image of second tunnel vision signal Difference；If there is x comparison pel, x difference can be determined；Then, if difference is zero, it is determined that ratio corresponding with the difference Pel is consistent；If the consistent number for comparing pel, then can be within preset range for a current frame image The the i-th frame image for determining first via vision signal and the jth frame image of the second tunnel vision signal are consistent, and otherwise determine first I-th frame image of road vision signal and the jth frame image of the second tunnel vision signal are inconsistent.
The present embodiment by executing following mistake for first via vision signal to be compared and the second tunnel vision signal respectively Journey: present frame figure is determined according to the image feature value of each pixel of current frame image for each frame image of vision signal The fit characteristic value of each comparison pel of picture；Wherein, respectively compare pel be by a frame image be divided into predetermined number region and It obtains；The fit characteristic value for determining each comparison pel of the first via vision signal on each frame image, with the second road video Whether the fit characteristic value of each comparison pel of the signal on each frame image is identical, to find first via vision signal and second The video frame synchronization point of road vision signal, wherein video frame synchronization point characterizes first via vision signal and the second tunnel vision signal It is initially synchronous from the video frame synchronization point；For each frame image since video frame synchronization point, determine that the first via regards The fit characteristic value of each comparison pel of the frequency signal on current frame image is believed with the second tunnel vision signal with first via video Number current frame image corresponding to each comparison pel on image fit characteristic value between each difference, and according to each difference Determine whether first via vision signal and the second tunnel vision signal are consistent on each frame image.To provide a kind of new view The comparison that two-path video signal carries out vision signal is analyzed, can more accurately monitor to need to carry out by frequency comparison method Whether abnormal broadcast vision signal；Also, the division that region is carried out for each frame image of vision signal, obtains at least one Pel is compared, the consistency of each frame image is analyzed using the fit characteristic value for comparing pel, the video analysis provided Method accurate and effective, can effectively identify whether each frame image of vision signal has occurred exception, and then effectively know Not Chu vision signal whether be abnormal.
Fig. 2 is the flow chart one for the processing method that video provided by Embodiment 2 of the present invention compares, as shown in Fig. 2, this reality The method for applying example, comprising:
Step 201, for first via vision signal to be compared and the second tunnel vision signal, execute following procedure respectively: It is for each frame image of vision signal, Y, U, V SD of current frame image is special when vision signal is SD vision signal Value indicative is converted to R, G, B characteristic value, and according to preset compensation factor, R, G, B characteristic value are converted to Y, U, V high definition feature Value；Wherein, Y, U, V high definition characteristic value are the image feature value of each pixel of current frame image.
Wherein, the Y high definition characteristic value in image feature value is 0.21*R+0.71*G+0.07*B, the U in image feature value High definition characteristic value is (0.5*B-0.11*R-0.38G) * 1.02+128, and the V high definition characteristic value in image feature value is (0.5*R- 0.45*G-0.045B)*1.02+128。
In the present embodiment, specifically, based on digital component serial line interface (Serial Digital Interface, letter Claim SDI) high-definition video signal, have differences, needing high definition in color gamut space with the SD vision signal based on SDI When vision signal and SD vision signal are compared, need to carry out Y, U, V characteristic value of SD vision signal Chrominance space compensation.
In the present embodiment, main road vision signal can be known as to first via vision signal, standby road vision signal is known as Second tunnel vision signal；Alternatively, first via vision signal, the second tunnel vision signal are all main road vision signal or the first via Vision signal, the second tunnel road vision signal Dou Shibei vision signal.For example, first via vision signal is main road HD video Signal, the second tunnel vision signal are standby road high-definition video signal；Alternatively, first via vision signal is the clear vision signal of main road sign, Second tunnel vision signal is the standby clear vision signal of road sign；Alternatively, first via vision signal is main road high-definition video signal, the second tunnel Vision signal is the clear vision signal of main road sign；Alternatively, first via vision signal is standby road high-definition video signal, the second road video letter Number for the standby clear vision signal of road sign.
When there is SD vision signal in first via vision signal, the second tunnel vision signal, need to regard SD Frequency signal carries out chrominance space compensation.Firstly, each pixel of each frame image for SD vision signal, present frame figure The image feature value of the current pixel point of picture is Y, U, V high definition characteristic value, and Y, U, V of the current pixel point of current frame image are marked Clear characteristic value is respectively converted into R, G, B characteristic value, for example, R=Y+1.36* (V-128), G=Y-0.33* (U-128-0.7* (V-128), B=Y+1.72* (U-128)；Then, for each pixel of each frame image of SD vision signal, according to R, G, B characteristic value of the current pixel point of current frame image are converted to Y, U, V high definition feature by pre-set compensation factor Value, for example, Y high definition characteristic value is 0.21*R+0.71*G+0.07*B, U high definition characteristic value is (0.5*B-0.11*R-0.38*G) * 1.02+128, V high definition characteristic value are (0.5*R-0.45*G-0.045*B) * 1.02+128.Y, U, the V used in the next steps Characteristic value, for Y, U, V high definition characteristic value here.
Y, U, V high definition characteristic value can also be known as to Y compensation, U compensation, V compensation, Y, U, V high definition characteristic value are respectively pair Y, U, V characteristic value of SD vision signal compensate after Y, U, V characteristic value；Y, U, V of SD video just may be used after being compensated It is calculated with carrying out the vector of subsequent step.
Step 202, for first via vision signal to be compared and the second tunnel vision signal, execute following procedure respectively: Current frame image is determined according to the image feature value of each pixel of current frame image for each frame image of vision signal Each comparison pel fit characteristic value；Wherein, respectively comparing pel is that a frame image is divided into the region of predetermined number and is obtained It arrives；
Wherein, step 202, it specifically includes:
For first via vision signal to be compared and the second tunnel vision signal, following procedure is executed respectively:
For each frame image, by the pixel of every X adjacent position of current frame image, it is determined as a basis Pel, wherein X is positive integer；
It is special according to the image of each pixel in the basis pel for each basic pel of each frame image Value indicative determines the foundation characteristic value of the basis pel；
For each frame image, by the basic pel per N number of adjacent position of current frame image, it is determined as a ratio To pel, wherein N is positive integer；
For each comparison pel of each frame image, according to the basis of each basic pel in the comparison pel Characteristic value determines the vector characteristic value of the comparison pel；
For each comparison pel of each frame image, according to the vector characteristic value of the comparison pel, determining should Compare the fit characteristic value of pel.
It wherein, include Y characteristic value, U characteristic value and V characteristic value in image feature value；It include Y spy in foundation characteristic value Levy mean value, U characteristic mean and V characteristic mean；It include Y ' characteristic value, U ' characteristic value and V ' characteristic value in vector characteristic value；It is quasi- Conjunction characteristic value is P=a*Y '+b*U '+c*V ', wherein a, b and c are weighting coefficient.
In the present embodiment, specifically, being required for first via vision signal to be compared and the second tunnel vision signal The fit characteristic value of each respective comparison pel is determined respectively.First via vision signal and the second road video can be believed Number, the process of this step is executed, respectively to determine that each of each frame image compares pel in the second tunnel vision signal Fit characteristic value.
It is described below and vision signal is handled, to determine each comparison chart of each frame image in vision signal The process of the fit characteristic value of member.
Firstly, carrying out the acquisition of vision signal, de-embedding is carried out to the vision signal based on SDI；Then, to vision signal into The preliminary signal processing of row, that is, carry out the extraction of the image feature value of each pixel, the characteristics of image of each pixel Value includes Y characteristic value, U characteristic value and V characteristic value.
Then, for each frame image of vision signal, by the pixel of every X adjacent position of current frame image Point is determined as a basic pel, wherein X is positive integer.For example, Fig. 3 is video ratio provided by Embodiment 2 of the present invention Pair processing method in basic pel schematic diagram, as shown in figure 3, for each frame image of vision signal, by every X phase The pixel that ortho position is set is determined as a basic pel, and then each frame image is divided into the basic pel of m*n block, wherein n Indicate that the line number of basic pel, m indicate the columns of basic pel.
Then, for each basic pel of each frame image of vision signal, according in current basal pel Each pixel image feature value, calculate the foundation characteristic value of current basal pel；Wherein, include in image feature value Y characteristic value, U characteristic value and V characteristic value；It include Y characteristic mean in foundation characteristic valueU characteristic meanWith V characteristic meanSpecifically, the mean value for seeking the Y characteristic value of each pixel in current basal pel, the Y as current basal pel are special Levy mean valueThe mean value for seeking the U characteristic value of each pixel in current basal pel, the U feature as current basal pel Mean valueThe mean value for seeking the V characteristic value of each pixel in current basal pel, the V feature as current basal pel are equal Value
Then, for each frame image of vision signal, by the foundation drawing per N number of adjacent position of current frame image Member is determined as a comparison pel, wherein N is positive integer.For example, Fig. 4 is video ratio provided by Embodiment 2 of the present invention Pair processing method in the schematic diagram one of comparison pel can be by every 4 adjacent bits as shown in figure 4, for each frame image The basic pel set is determined as a comparison pel, i.e., the area of basic pel is the 1/4 of the area of comparison pel；N ' expression ratio To the line number of pel, m ' expression compares the columns of pel.
Then, each of each frame image of vision signal is compared for pel, is compared in pel according to current Each basic pel foundation characteristic value, determine the current vector characteristic value for comparing pel；Wherein, include in vector characteristic value Y ' characteristic value, U ' characteristic value and V ' characteristic value.Specifically, by the current Y characteristic mean for comparing each basic pel in pel Radian value conversion is carried out, the Y ' characteristic value for currently comparing pel is obtained；By the current U feature for comparing each basic pel in pel Mean value carries out radian value conversion, obtains the U ' characteristic value for currently comparing pel；By the current V for comparing each basic pel in pel Characteristic mean carries out radian value conversion, obtains the V ' characteristic value for currently comparing pel.Wherein, the process of radian value conversion is a kind of The process of Vectorization Algorithm.
For example, Fig. 5 is showing for the comparison pel in the processing method of video provided by Embodiment 2 of the present invention comparison It is intended to two, as shown in figure 5, this 4 basic pels constitute a comparison pel for the basic pel of every 4 adjacent positions, Per adjacent 4, the Y characteristic mean of basic pel is respectivelyIt then will by algorithm Y ' the characteristic value of this 4 characteristic mean resultant vector characteristic values, can be by the Y characteristic mean of corresponding 4 basic pels in the side x It is up-converted to radian value to, the side y, obtaining Y ' characteristic value isThe basic pel per adjacent 4 U characteristic mean is respectivelyThen this 4 characteristic mean synthesis are sweared by algorithm The Y characteristic mean of corresponding 4 basic pels can be up-converted to arc in the direction x, the side y by the U ' characteristic value of measure feature value Angle value, obtaining U ' characteristic value isPer adjacent 4, the V characteristic mean of basic pel is respectivelyThen pass through algorithm for the U ' feature of this 4 characteristic mean resultant vector characteristic values The Y characteristic mean of corresponding 4 basic pels can be up-converted to radian value in the direction x, the side y, obtained by valueCharacteristic value For
Then, for for each comparison pel of each frame image, in the current vector characteristic value for comparing pel Y ' characteristic value, U ' characteristic value and V ' characteristic value be weighted after processing, obtain currently compare pel fit characteristic value；Its In, fit characteristic value is P=a*Y '+b*U '+c*V ', and a, b and c are weighting coefficient, and Y ' characteristic value, U ' characteristic value and V ' feature Value is vector characteristic value.One comparison pel has a fit characteristic value.
Step 203, the fit characteristic value for determining each comparison pel of the first via vision signal on each frame image, with Whether the fit characteristic value of each comparison pel of the two tunnel vision signals on each frame image is identical, to find first via video letter Video frame synchronization point number with the second tunnel vision signal, wherein video frame synchronization point characterizes first via vision signal and the second tunnel Vision signal from the video frame synchronization point be initially synchronous.
Wherein, step 203, it specifically includes:
The fit characteristic value for determining each comparison pel of the first via vision signal on each frame image, with the second road video Whether the fit characteristic value of each comparison pel of the signal on each frame image is identical, to determine first via vision signal and second Whether each comparison pel of the road vision signal on each frame image be identical, and determines first via vision signal and the second road video Signal compares the different block number of pel on each frame image；
Determining first via vision signal and the second tunnel vision signal comparison chart in each frame image of continuous P frame image Member different block number when less than the first block number threshold value, determines that the frame after P frame is video frame synchronization point；Alternatively, determining The vision signal block number that from the second tunnel vision signal to compare pel in each frame image of continuous Q frame image different all the way is small When the second block number threshold value, determine that the frame after Q frame is video frame synchronization point；
Wherein, P, Q are positive integer, and P is greater than Q, and the first block number threshold value is less than the second block number threshold value.
In the present embodiment, specifically, next needing to find the view of first via vision signal Yu the second tunnel vision signal Frequency frame synchronization point.Firstly the need of the fit characteristic value of each comparison pel by first via vision signal on each frame image, with The fit characteristic value of each comparison pel of the second tunnel vision signal on each frame image is compared, be confirmed whether it is identical, into And determine first via vision signal and the second tunnel vision signal on each frame image whether respectively compare pel identical, for example, By the first frame image of first via vision signal, the second frame image ..., N1 frame image, first with the second tunnel vision signal Frame image, the second frame image ..., N2 frame image is compared analysis；By each of the first frame image of first via vision signal Compare pel fit characteristic value, respectively with the fit characteristic for respectively comparing pel of the first frame image of the second tunnel vision signal Value, the second frame image each comparison pel fit characteristic value ..., the fit characteristic value of each comparison pel of N2 frame image into Row compares；By by the fit characteristic value of each comparison pel of the second frame image of first via vision signal, regarded respectively with the second tunnel The fit characteristic of the fit characteristic value of each comparison pel of the first frame image of frequency signal, each comparison pel of the second frame image Value ..., the fit characteristic value of each comparison pel of N2 frame image is compared；And so on.
It is being determined whether first via vision signal and the second tunnel vision signal respectively compare pel on each frame image After identical, the first via vision signal block different from comparison pel of the second tunnel vision signal on each frame image is counted Number.
Then, determining first via vision signal and the second tunnel vision signal in each frame image of continuous P frame image The different block number of pel is compared, when less than the first block number threshold value, so that it may determine that the frame after P frame is video frame synchronization point； Alternatively, determining first via vision signal and the second tunnel vision signal compares pel in each frame image of continuous Q frame image Different block number, when less than the second block number threshold value, so that it may determine that the frame after Q frame is video frame synchronization point；Wherein, P, Q For positive integer, P is greater than Q, and the first block number threshold value is less than the second block number threshold value.
For example, determining first via vision signal and the second tunnel vision signal in each frame figure of continuous 30 frame image The different block number of pel is compared as in, when less than 15, so that it may determine that the frame after 30 frames is video frame synchronization point；Alternatively, It is different to determine that first via vision signal from the second tunnel vision signal compares pel in each frame image of continuous 15 frame image Block number, when less than 40, so that it may determine that the frame after 40 frames is video frame synchronization point.
Specifically, the specific strategy of the video frame synchronization point of first via vision signal and the second tunnel vision signal is found such as Lower process.First against first via vision signal, the second tunnel vision signal, a FIFO buffer area is established respectively and carries out data Caching, is respectively defined as A buffer area and B buffer area.
First need to set synchronous condition as first via vision signal is with the second tunnel vision signal in continuous P 1 or Q1 frame phase Together.And since there are actual frame difference S between first via vision signal and the second tunnel vision signal, and then need in view of actual frame The problem of difference；It is to which synchronous condition finally be arranged, first via vision signal and the second tunnel vision signal are in continuous P=P1+S Or Q=Q1+S frame is identical.For example, S value can be 15 frames, and the value upper limit of S can be set as 750 frames, i.e., 30 seconds.From And the size of FIFO buffer area is P or Q, and P=P1+S, Q=Q1+S.It is below P with the size of FIFO buffer area, is situated between It continues.
It arrives in the 1st frame image a [1] of first via vision signal and the 1st frame image b [1] of the second tunnel vision signal When, respectively there are the data of 1 frame image in A buffer area and B buffer area.Then, start the comparison knot of statistics a [1] and b [1] Fruit.Specifically, judge that the fitting of the fit characteristic value of the 1st comparison pel of a [1] and the 1st comparison pel of b [1] is special Whether value indicative is identical, and if they are the same, then the 1st comparison pel that the 1st of a [1] compares pel and b [1] is identical, if not phase Together, then the 1st comparison pel that the 1st of a [1] compares pel and b [1] is not identical；Equally, judge the 2nd ratio of a [1] It is whether identical to the fit characteristic value of the 2nd comparison pel of the fit characteristic value of pel and b [1], if they are the same, then the of a [1] 2 the 2nd comparisons pels for comparing pels and b [1] be it is identical, if not identical, the 2nd comparison pel of a [1] and b [1] 2nd comparison pel is not identical；And so on, until judge a [1] the last one compare pel fit characteristic value with Whether the fit characteristic value that the last one of b [1] compares pel is identical, to judge the last one comparison pel and b of a [1] Whether [1] the last one compares pel identical；Then count compared in a [1] and b [1] the different block number of pel and, should Whether block number has obtained the comparison result of a [1] and b [1] less than the first block number threshold value.By the comparison of a [1] and b [1] As a result it is recorded in RA [0] and RB [0], respectively there is the comparison result of 1 pair of data in RA [0] and RB [0].
It arrives in the 2nd frame image a [2] of first via vision signal and the 2nd frame image b [2] of the second tunnel vision signal When, respectively there are the data of 2 frame images in A buffer area and B buffer area；Then the comparison result of a [2] and b [2], a are counted [2] identical with the comparison process of b [1] with a [1] with the comparison process of b [2], it is different with pel is compared in b [2] to count a [2] Block number and the block number whether less than the first block number threshold value, and then obtained the comparison result of a [2] and b [2].By a [2] It is not recorded in RA [0] and RB [0] with the comparison result of b [2], at this point, respectively there is the comparison knot of 2 pairs of data in RA [0] and RB [0] Fruit；And the comparison result of a [1] and b [2] are recorded in RA [1], at this point, RA by the comparison result for counting a [1] and b [2] [1] there is the comparison result of 1 pair of data in；And the comparison result for counting a [2] and b [1], by the comparison result of a [2] and b [1] It is recorded in RB [1], at this point, there is the comparison result of 1 pair of data in RB [1].
It arrives in the 3rd frame image a [3] of first via vision signal and the 3rd frame image b [3] of the second tunnel vision signal When, respectively there are the data of 3 frame images in A buffer area and B buffer area；Then the comparison result for counting a [3] and b [3], by a [3] and the comparison result of b [3], it is recorded in RA [0] and RB [0], respectively there is the comparison of 3 pairs of data in RA [0] and RB [0] at this time As a result；The comparison result for counting a [2] and b [3], the comparison result of a [2] and b [3] is recorded in RA [1], at this point, RA [1] In have the comparison results of 2 pairs of data；The comparison result for counting a [3] and b [2], is recorded in RB for the comparison result of a [3] and b [2] [1] in, at this point, RB [1] has the comparison result of 2 pairs of data；The comparison result for counting a [1] and b [3], by the ratio of a [1] and b [3] Result is recorded in RA [2], at this point, having the comparison result of 1 pair of data in RA [2]；The comparison result of a [3] and b [1] is counted, The comparison result of a [3] and b [1] are recorded in RB [2], there is the comparison result of 1 pair of data in RB [2] at this time.
And so on, in the P1 frame image a [P1] of first via vision signal and the P1 frame of the second tunnel vision signal When image b [P1] arrives, respectively there are the data of P1 frame image in A buffer area and B buffer area.Count a [P1] and b [P1] The comparison result of a [P1] and b [P1] are recorded in RA [0] and RB [0] by comparison result, at this point, respectively having in RA [0] and RB [0] Comparison result of the P1 to data.Judge that RA [0] meets synchronous thresholding this when, i.e. in the frame image of P1 in total in RA [0], The first via vision signal block number that from the second tunnel vision signal to compare pel in each frame image of 1 frame image of continuous P different Less than the first block number threshold value；If RA [0] meets synchronous thresholding, RA [0] at this time is recorded as RA_best, if RB [0] is accorded with Contract walks thresholding, and RB [0] at this time is recorded as RB_best.It is similar with the comparison process of each frame before, it is also necessary to count a The comparison result of a [P1-1] and b [P1] are recorded in RA [1] by the comparison result of [P1-1] and b [P1], at this point, in RA [1] There is P1-1 to the comparison result of data；The comparison result for counting a [P1] and b [P1-1], by the comparison result of [P1] and b [P1-1] It is recorded in RB [1], at this point, having in RB [1] P1-1 to the comparison result of data；Until counting on the comparison of a [1] and b [P1] As a result, the comparison result of a [1] and b [P1] are recorded in RA [P1-1], at this point, there is the comparison knot of 1 pair of data in RA [P1-1] Fruit；The comparison result for counting a [P1] and b [1], the comparison result of a [P1] and b [1] is recorded in RB [P1-1], at this point, RB There is the comparison result of 1 pair of data in [P1-1].
Then, in the P1+1 frame image a [P1+1] of the first via vision signal and P1+1 of the second tunnel vision signal When frame image b [P1+1] arrives, respectively there are the data of P1+1 frame image in A buffer area and B buffer area.It needs this when The comparison result of a [1] and b [1] is deleted from the RA [0] and RB [0], and increase a [P1+1] and b [comparison result of [P1+1], Respectively there is P1 to the comparison result of data in RA [0] and RB [0] at this time；Meet synchronous thresholding when determining [0] RA at this time, and excellent When RA_best before, then with the RA_best before the covering of RA [0] at this time, new RA_best is obtained；Determining this When RB [0] meet synchronous thresholding, and when RB_best before being better than, then with the RB_ before the covering of RB [0] at this time Best obtains new RB_best.Then, the comparison result of a [1] and b [2] is deleted from RA [1], and increases a [P1-1] and b The comparison result of [P1], RA [1] respectively has P1 to the comparison result of data at this time；It is determining that RA [1] meets synchronous thresholding, and is being better than The RA_best obtained before, the then RA_best obtained before with RA [1] covering at this time, then obtain a new RA_best. Comparison result from deletion a [2] and b [1] in RB [1], and increase the comparison result of a [P1] and b [P1-1], RB [1] is each at this time There is P1 to the comparison result of data；Equally, judge if meeting synchronous thresholding when to RB [1].And so on, it counts a [2] It with the comparison result of b [P1+1], is recorded in RA [P1-1], there is the comparison result of 2 pairs of data in RA [P1-1] at this time；Count a The comparison result of [P1+1] and b [2] are recorded in RB [P1-1], at this point, there is the comparison result of 2 pairs of data in RB [P1-1].
Wherein, the RB_best before RA [p] and RA_best before being better than or RB [p] are better than, p be 0 to P-1 it Between integer, here the different block number of pel is compared in each frame image better than referring in current RA [p], be less than it The different block number of pel is compared in preceding RA_best in each frame image, comparison chart in each frame image in RB current in other words [p] The different block number of member compares the different block number of pel in each frame image in the RB_best before being less than.For example, RA_ Have the comparison result of 3 pairs of data in best, a [1] and the comparison result of b [1] be compare the different block number of pel be 20, a [2] and The comparison result of b [2] is that compare the different block number of pel be 10, a [3] and the comparison result of b [3] block that be comparison pel different Number is 15；There is the comparison result of 3 pairs of data in RA [1], a [1] is to compare the different block number of pel to be with the comparison result of b [2] 5, a [2] and the comparison result of b [3] are to compare that the different block number of pel is 8, a [3] and the comparison result of b [4] is comparison pel Different block numbers is 5；It is found that the different total block data of the comparison pel of each frame image in RA [1], is less than each frame in RA_best Image compares the different total block data of pel, can determine the RA_best before RA [1] is better than, at this moment video frame synchronization point is 4th frame of first via vision signal and the 5th frame of the second tunnel vision signal.
It is similar with the analytic process of P1+1 frame image, carry out the judgement of subsequent each frame.In first via vision signal When P=P1+S frame image a [P] and the P=P1+S frame image b [P] of the second tunnel vision signal arrives, cached in A Respectively there are the data of P frame image in area and B buffer area.From the comparison knot for deleting a [1] and b [1] in RA [P-P1] and RB [P-P1] Fruit, and increase the comparison result of a [N] and b [N], it is separately recorded in RA [0] and RB [0], respectively has P1 to the comparison knot of data Fruit；If RA [0] meets synchronous thresholding, and better than newest RA_best obtained in the above process, is then covered with RA [0] at this time The RA_best obtains a new RA_best；If RB [0] meets synchronous thresholding, and newest better than obtained in the above process RB_best, then cover the RB_best with RB at this time [0], obtain a new RB_best.And so on.Then from RA [1] The middle comparison result for deleting a [1] and b [S], and increase the comparison result of a [P-S-1] and b [P], RA [S-1] respectively has M pairs at this time The comparison result of data；If RA [S-1] meets synchronous thresholding, and is better than newest RA_best, then covered with RA [S-1] at this time Newest RA_best obtains a new RA_best.From the comparison result of deletion a [S] and b [1] in RB [1], and increase a The comparison result of [P] and b [P-S-1], RB [1] respectively has M to the comparison result of data at this time；If RB [S-1] meets synchronous thresholding, And be better than newest RB_best, then newest RB_best is covered with RB [S-1] at this time, obtains a new RB_best.
It is similar with above procedure, it can be for first via vision signal and the second tunnel vision signal in continuous Q=Q1+S frame Upper identical condition, is analyzed.Similar, comparison result is recorded in CRA [0]~CRA [S-1] and CRB [0]~CRB [S- 1] in.CRA is similar with the statistical method of RA, RB with the statistics of CRB, and details are not described herein again.Wherein, if CRA [0]~CRA [S-1] In have meet compare thresholding as a result, it is true for then defining bFindNoramal_A；If having in CRB [0]~CRB [S-1] and meeting ratio To thresholding as a result, it is true for then defining bFindNoramal_B.
Then, it needs to be compared analysis to finally obtained RA [0] and RA [0], if finally obtained RA [0] and RA [0] result is consistent, then needs to judge to be currently found whether video frame synchronization point is still frame, Hei Chang, color field, one in colour bar Kind；If it is one of still frame, Hei Chang, color field, colour bar that video frame synchronization point, which is currently found, does not generate and compare abnormal report It is alert, and relative recording is resetted, video frame synchronization point is looked for again；If it is not still frame, Hei Chang, coloured silk that video frame synchronization point, which is currently found, One of field, colour bar can then determine that RB_best is final result.
Fig. 6 is the process of the searching video frame synchronization point in the processing method that video provided by Embodiment 2 of the present invention compares Figure, as shown in fig. 6, including step 301- step 308, main process is as follows:
Step 301 determines whether to have found video frame synchronization point.
Step 302 determines whether to synchronize success.
Step 303, after step 302, if not synchronizing success, it is determined that RA_best and RB_best.
In this step, it after success synchronous with the second tunnel vision signal not by first via vision signal, still needs It to go to analyze RA_best and RB_best using above procedure.
Step 304, after step 303 judges whether to reach the thresholding frame number found and judge synchronized result.
In step 304, it refers here to be necessary to determine whether to analyze P frame image or Q frame image.If step Confirmation reaches thresholding after rapid 304, thens follow the steps 301.
Step 305 judges whether to define video frame synchronization point.
In this step, it if finally obtained RA_best and RB_best exist, needs to finally obtained RA_best It is compared with RB_best, takes conduct synchronized result optimal in the two；Specifically, each frame in current RA_best The different block number of pel is compared in image, if the block number different less than comparison pel in frame image each in current RB_best； If being less than, it is determined that RA_best is optimal, however, it is determined that is more than or equal to, it is determined that RB_best is optimal.If RA_best is deposited , but RB_best is not present, and takes RA_best as final result；If RA_best is not present, RB_best exists, and takes RB_ Best is as final result.
If RA_best and RB_best are not present, need to judge whether bFindNoramal_A or BFindNoramal_B is true；If it is determined that bFindNoramal_A or bFindNoramal_B are very, then it is abnormal not generate comparison Alarm, and reset relative recording, look for video frame synchronization point again；If it is determined that bFindNoramal_A and bFindNoramal_B It is not very, then to generate and compare inconsistent information, and reset relative recording, look for video frame synchronization point again.
Step 306, if it is determined that video frame synchronization point, then carry out first via vision signal and the second tunnel vision signal It is synchronous.
In this step, if after step 306, if needing to find video frame synchronization point again thens follow the steps 301.
If step 307 is not determined by video frame synchronization point, relative recording is resetted, finds video frame synchronization point again.
In this step, if executing step 301 after step 307.
Step 308, after step 302, if synchronizing success, to first via vision signal and the second tunnel vision signal into On each frame image of row whether consistent analysis.
After step 203, step 204 is executed.
Step 204, for each frame image since video frame synchronization point, determine first via vision signal in present frame The fit characteristic value of each comparison pel on image, with the second tunnel vision signal in the current frame image with first via vision signal Each difference between the fit characteristic value of each comparison pel on corresponding image, and first via video is determined according to each difference Whether signal and the second tunnel vision signal are consistent on each frame image.
Wherein, step 204, it specifically includes:
Each frame image since video frame synchronization point determines for each comparison pel of each frame image The fitting of the current comparison pel of the fit characteristic value and the second tunnel vision signal of the current comparison pel of first via vision signal Difference between characteristic value；
Each frame image since video frame synchronization point, for each comparison pel of each frame image, true When determining difference less than or equal to difference threshold, the current comparison pel and the second tunnel vision signal of first via vision signal are determined Current comparison pel be consistent；When determining that difference is greater than difference threshold, the current ratio of first via vision signal is determined Current comparison pel to pel and the second tunnel vision signal is inconsistent, and it is different that pel is compared in current frame image The block number of cause；
Each frame image since video frame synchronization point is determining that comparison pel is different for each frame image When the block number of cause is less than third block number threshold value, determine first via vision signal and the second tunnel vision signal on current frame image It is consistent；When determining that comparing the inconsistent block number of pel is more than or equal to third block number threshold value, determine that first via video is believed It number is inconsistent on current frame image with the second tunnel vision signal；
Each frame image since video frame synchronization point is determining first via vision signal and the second tunnel vision signal exists When being inconsistent on continuous Z frame image, it is abnormal to determine that video compares, and issues the abnormal information warning of comparison, wherein Z is positive whole Number；
After issuing the abnormal information warning of comparison, determining first via vision signal with the second tunnel vision signal continuous When being consistent on M frame image, it is normal to determine that video compares, wherein M is positive integer.
In the present embodiment, specifically, first via vision signal and the second tunnel vision signal are directed to, from video frame synchronization Each frame image that point starts calculates the current of first via vision signal for each comparison pel of each frame image Compare the difference between the fit characteristic value of pel and the fit characteristic value of the current comparison pel of the second tunnel vision signal；So After be less than or equal to difference threshold when judging the difference；If it is determined that the difference is less than or equal to difference threshold, it is determined that first The current comparison pel of road vision signal and the current comparison pel of the second tunnel vision signal are consistent；If it is determined that difference is big In difference threshold, it is determined that the current comparison pel of first via vision signal and the current comparison chart of the second tunnel vision signal Member is inconsistent.For each frame image since video frame synchronization point, in each ratio for having counted current frame image After whether consistent to pel, need to count the block number that comparison pel is inconsistent in current frame image.
Then, judge whether the inconsistent block number of comparison pel of current frame image is less than third block number threshold value；If really The inconsistent block number of pel that compares of settled prior image frame is less than third block number threshold value, it is determined that first via vision signal and the Two tunnel vision signals are consistent on current frame image；If it is determined that comparing the inconsistent block number of pel is more than or equal to third block number When threshold value, it is determined that first via vision signal and the second tunnel vision signal are inconsistent on current frame image.
For example, video frame synchronization point indicates the i-th frame image from first via vision signal, the second tunnel vision signal Jth frame image starts, and first via vision signal is synchronous with the second tunnel vision signal；It can so believe from first via video Number the i-th frame image start to analyze first via vision signal, to since the jth frame image of the second tunnel vision signal Two tunnel vision signals are analyzed.
The i-th frame image of first via vision signal is compared with the jth frame image of the second tunnel vision signal first point Analysis；The fit characteristic value for first calculating the 1st comparison pel of the i-th frame image of first via vision signal, with the second road video Difference between the fit characteristic value of 1st comparison pel of the jth frame image of signal, if the difference is less than or equal to difference threshold Value, it is determined that the 1st comparison pel of the i-th frame image of first via vision signal, the jth frame image with the second tunnel vision signal The 1st comparison pel be consistent, if the difference be greater than difference threshold, it is determined that the i-th frame figure of first via vision signal 1st comparison pel of picture, it is inconsistent for comparing pel with the 1st of the jth frame image of the second tunnel vision signal；With such It pushes away, compares the last one comparison pel of the i-th frame image of first via vision signal, the jth frame with the second tunnel vision signal The last one of image compares whether pel is consistent, and the i-th frame image and the second tunnel for then counting first via vision signal regard The jth frame image of frequency signal compares the inconsistent block number of pel；Then, if the block number is less than third block number threshold value, really The the i-th frame image for determining first via vision signal and the jth frame image of the second tunnel vision signal are consistent, if the block number be greater than etc. In third block number threshold value, it is determined that the i-th frame image of first via vision signal is with the jth frame image of the second tunnel vision signal Inconsistent.
And so on ,+1 frame image of jth of i+1 frame image and the second tunnel vision signal to first via vision signal into Row compare analysis, determine first via vision signal i+1 frame image and the second tunnel vision signal+1 frame image of jth whether It is consistent.
By analyzing frame by frame, if detecting, first via vision signal is on continuous Z frame image with the second tunnel vision signal When inconsistent, wherein Z is positive integer, so that it may determine that current video comparison result is that video compares exception, and puts down to management Platform, which is sent, compares abnormal information warning.Then, after issuing the comparison exception information warning, if occurring the determining first via again It is normal can to determine that video compares when being consistent situation on continuous N frame image for vision signal and the second tunnel vision signal, this When not alert, wherein M is positive integer.
Fig. 7 is the flowchart 2 for the processing method that video provided by Embodiment 2 of the present invention compares, as shown in fig. 7, comprises:
Step 401, initialization.
Step 402, the video frame synchronization point for determining first via vision signal Yu the second tunnel vision signal.
Step 403 judges the analysis result of video frame synchronization point, and determines whether successfully to search out video frame same Beans-and bullets shooter.
In this step, step 402-403 can execute step 201,202, the process of step 203.
If step 404 finds the failure of video frame synchronization point, warning message, the warning message table are sent to network management platform Levy audio video synchronization failure.
If step 405 finds the success of video frame synchronization point, synchronous successful information is sent to network management platform.
Step 406, the record first via vision signal frame difference relationship synchronous with the generation of the second tunnel vision signal.
In this step, if frame difference relationship refers to that video frame synchronization point indicates the i-th frame figure from first via vision signal The jth frame image of picture, the second tunnel vision signal starts, and first via vision signal is synchronous with the second tunnel vision signal；So First via vision signal can be analyzed since the i-th frame image of first via vision signal, be believed from the second road video Number jth frame image start to analyze the second tunnel vision signal.
Frame difference relationship is sent to audio comparison module by step 407.
Step 408, the comparison point according to frame difference relationship, to first via vision signal and the second tunnel vision signal progress image Analysis.
In this step, step 408, the process of step 204 can be executed.
Step 409, the comparison result for determining first via vision signal Yu the second tunnel vision signal, if compared just for video Often.Here, if the result of video comparison is identical as last comparison result, 408 are thened follow the steps.
Step 410, if it is determined that video compare it is normal and different from last comparison result, then to network management platform transmission ratio Request to abnormal restoring, and execute step 408.
Step 411, if it is determined that video compare it is abnormal and different from last comparison result, it is determined whether progress video Fast synchronization.Here, if the Fast synchronization success of video, thens follow the steps 406.
If the Fast synchronization of step 412, video fails, is sent to network management platform and compare abnormal message, and execute step Rapid 408.
The present embodiment by executing following mistake for first via vision signal to be compared and the second tunnel vision signal respectively Journey: present frame figure is determined according to the image feature value of each pixel of current frame image for each frame image of vision signal The fit characteristic value of each comparison pel of picture；Wherein, respectively compare pel be by a frame image be divided into predetermined number region and It obtains；The fit characteristic value for determining each comparison pel of the first via vision signal on each frame image, with the second road video Whether the fit characteristic value of each comparison pel of the signal on each frame image is identical, to find first via vision signal and second The video frame synchronization point of road vision signal, wherein video frame synchronization point characterizes first via vision signal and the second tunnel vision signal It is initially synchronous from the video frame synchronization point；For each frame image since video frame synchronization point, determine that the first via regards The fit characteristic value of each comparison pel of the frequency signal on current frame image is believed with the second tunnel vision signal with first via video Number current frame image corresponding to each comparison pel on image fit characteristic value between each difference, and according to each difference Determine whether first via vision signal and the second tunnel vision signal are consistent on each frame image.Wherein, it is wrapped in image feature value Y characteristic value, U characteristic value and V characteristic value are included.To provide a kind of new video comparison method, by two-path video signal into The comparison of row vision signal is analyzed, and whether extremely can more accurately monitor broadcast vision signal；Also, needle The division that region is carried out to each frame image of vision signal obtains at least one and compares pel, using the fitting for comparing pel Characteristic value analyzes the consistency of each frame image, and the video analysis method accurate and effective provided can be identified effectively Whether each frame image of vision signal has occurred exception out, and then effectively identifies whether vision signal is abnormal；And And it is mainly based upon what Y characteristic value was compared compared to existing video alignments, and since video is in encoding and decoding, biography The complexity of defeated equal links process, the comparison based on Y characteristic value, which often exists, compares wrong report, in this application to Y characteristic value, U Characteristic value and V characteristic value are analyzed, and vector characteristic value is obtained, then obtain a fit characteristic value, and then can be special based on YUV Value indicative carries out analysis comparison to video, can effectively reduce wrong report, improves and compares accuracy.
Fig. 8 is the structural schematic diagram for the processing unit that the video that the embodiment of the present invention three provides compares, as shown in figure 8, this The device of embodiment, comprising:
Determining module 81, for for first via vision signal to be compared and the second tunnel vision signal, execute respectively with Lower process: it is determined current for each frame image of vision signal according to the image feature value of each pixel of current frame image The fit characteristic value of each comparison pel of frame image；Wherein, respectively comparing pel is the area that a frame image is divided into predetermined number Obtained from domain；
Analysis module 82, for determining the fit characteristic of each comparison pel of the first via vision signal on each frame image Value, it is whether identical as the fit characteristic value that respectively compares pel of the second tunnel vision signal on each frame image, to find first The video frame synchronization point of road vision signal and the second tunnel vision signal, wherein video frame synchronization point characterizes first via vision signal With the second tunnel vision signal from the video frame synchronization point be initially synchronous；
Comparison module 83, for determining first via vision signal for each frame image since video frame synchronization point The fit characteristic value of each comparison pel on current frame image is worked as with the second tunnel vision signal with first via vision signal Each difference between the fit characteristic value of each comparison pel on image corresponding to prior image frame, and is determined according to each difference Whether vision signal and the second tunnel vision signal are consistent on each frame image all the way.
The processing that the video that the embodiment of the present invention one provides compares can be performed in the processing unit that the video of the present embodiment compares Method, realization principle is similar, and details are not described herein again.
The present embodiment by executing following mistake for first via vision signal to be compared and the second tunnel vision signal respectively Journey: present frame figure is determined according to the image feature value of each pixel of current frame image for each frame image of vision signal The fit characteristic value of each comparison pel of picture；Wherein, respectively compare pel be by a frame image be divided into predetermined number region and It obtains；The fit characteristic value for determining each comparison pel of the first via vision signal on each frame image, with the second road video Whether the fit characteristic value of each comparison pel of the signal on each frame image is identical, to find first via vision signal and second The video frame synchronization point of road vision signal, wherein video frame synchronization point characterizes first via vision signal and the second tunnel vision signal It is initially synchronous from the video frame synchronization point；For each frame image since video frame synchronization point, determine that the first via regards The fit characteristic value of each comparison pel of the frequency signal on current frame image is believed with the second tunnel vision signal with first via video Number current frame image corresponding to each comparison pel on image fit characteristic value between each difference, and according to each difference Determine whether first via vision signal and the second tunnel vision signal are consistent on each frame image.To provide a kind of new view The comparison that two-path video signal carries out vision signal is analyzed, can more accurately monitor to need to carry out by frequency comparison method Whether abnormal broadcast vision signal；Also, the division that region is carried out for each frame image of vision signal, obtains at least one Pel is compared, the consistency of each frame image is analyzed using the fit characteristic value for comparing pel, the video analysis provided Method accurate and effective, can effectively identify whether each frame image of vision signal has occurred exception, and then effectively know Not Chu vision signal whether be abnormal.
Fig. 9 is the structural schematic diagram for the processing unit that the video that the embodiment of the present invention four provides compares, in embodiment three On the basis of, as shown in figure 9, the device of the present embodiment, determining module 81, comprising:
First determines submodule 811, for being directed to for each frame image, by every X adjacent position of current frame image Pixel, be determined as a basic pel, wherein X is positive integer；
Second determines submodule 812, for each basic pel for each frame image, according to the foundation drawing The image feature value of each pixel in member determines the foundation characteristic value of the basis pel；
Third determines submodule 813, for being directed to for each frame image, by current frame image per N number of adjacent position Basic pel, be determined as a comparison pel, wherein N is positive integer；
4th determines submodule 814, for comparing for pel for each of each frame image, according to the comparison chart The foundation characteristic value of each basic pel in member, determines the vector characteristic value of the comparison pel；
5th determines submodule 815, for comparing for pel for each of each frame image, according to the comparison chart The vector characteristic value of member, determines the fit characteristic value of the comparison pel.
It include Y characteristic value, U characteristic value and V characteristic value in image feature value；It include that Y feature is equal in foundation characteristic value Value, U characteristic mean and V characteristic mean；Second determines submodule 812, is specifically used for:
For each basic pel of each frame image, the Y feature of each pixel in the basis pel is determined The mean value of value, the mean value of the U characteristic value of each pixel, the mean value of the V characteristic value of each pixel, the respectively Y of the basis pel Characteristic mean, U characteristic mean, V characteristic mean.
It include Y ' characteristic value, U ' characteristic value and V ' characteristic value in vector characteristic value；4th determines submodule 814, specifically For:
For each comparison pel of each frame image, by the Y feature of each basic pel in the comparison pel Mean value carries out radian value conversion, obtains the Y ' characteristic value of the comparison pel, and the U of each basic pel in the comparison pel is special It levies mean value and carries out radian value conversion, obtain the U ' characteristic value of the comparison pel, and by the V of each basic pel in the comparison pel Characteristic mean carries out radian value conversion, obtains the V ' characteristic value of the comparison pel.
Fit characteristic value is P=a*Y '+b*U '+c*V ', wherein a, b and c are weighting coefficient.
Device provided in this embodiment, further includes:
Conversion module 91, for being directed to each frame image of vision signal in determining module 81, according to current frame image The image feature value of each pixel, before the fit characteristic value of each comparison pel for determining current frame image, for be compared First via vision signal and the second tunnel vision signal, execute following procedure respectively: when vision signal is SD vision signal, needle To each frame image of vision signal, by Y, U, V SD characteristic value of current frame image, R, G, B characteristic value are converted to, and according to R, G, B characteristic value are converted to Y, U, V high definition characteristic value by preset compensation factor；Wherein, Y, U, V high definition characteristic value are current The image feature value of each pixel of frame image.
Wherein, the Y high definition characteristic value in image feature value is 0.21*R+0.71*G+0.07*B, the U in image feature value High definition characteristic value is (0.5*B-0.11*R-0.38*G) * 1.02+128, and the V high definition characteristic value in image feature value is (0.5*R- 0.45*G-0.045*B)*1.02+128。
Analysis module 82, is specifically used for:
The fit characteristic value for determining each comparison pel of the first via vision signal on each frame image, with the second road video Whether the fit characteristic value of each comparison pel of the signal on each frame image is identical, to determine first via vision signal and second Whether each comparison pel of the road vision signal on each frame image be identical, and determines first via vision signal and the second road video Signal compares the different block number of pel on each frame image；
Determining first via vision signal and the second tunnel vision signal comparison chart in each frame image of continuous P frame image Member different block number when less than the first block number threshold value, determines that the frame after P frame is video frame synchronization point；Alternatively, determining The vision signal block number that from the second tunnel vision signal to compare pel in each frame image of continuous Q frame image different all the way is small When the second block number threshold value, determine that the frame after Q frame is video frame synchronization point；
Wherein, P, Q are positive integer, and P is greater than Q, and the first block number threshold value is less than the second block number threshold value.
Comparison module 83, comprising:
First compares submodule 831, for each frame image since video frame synchronization point, for each frame image For each comparison pel, the fit characteristic value and the second road video letter of the current comparison pel of first via vision signal are determined Number current comparison pel fit characteristic value between difference；
Second compares submodule 832, for each frame image since video frame synchronization point, for each frame image For each comparison pel, when determining that difference is less than or equal to difference threshold, the current comparison of first via vision signal is determined Pel and the current comparison pel of the second tunnel vision signal are consistent；When determining that difference is greater than difference threshold, the is determined The current comparison pel of vision signal and the current comparison pel of the second tunnel vision signal are inconsistent all the way, and current The inconsistent block number of pel is compared in frame image；
Third compares submodule 833, for each frame image since video frame synchronization point, for each frame image come It says, when determining that comparing the inconsistent block number of pel is less than third block number threshold value, determines first via vision signal and the second tunnel Vision signal is consistent on current frame image；Determining that comparing the inconsistent block number of pel is more than or equal to third block number thresholding When value, determine that first via vision signal and the second tunnel vision signal are inconsistent on current frame image；
First confirmation submodule 834 is determining first via video for each frame image since video frame synchronization point It is abnormal to determine that video compares when being inconsistent on continuous Z frame image for signal and the second tunnel vision signal, and it is abnormal to issue comparison Information warning, wherein Z is positive integer.
Second confirmation submodule 835 is used for after the first confirmation submodule 834 issues and compares abnormal information warning, First via vision signal and the second tunnel vision signal are determined when being consistent on continuous N frame image, it is normal to determine that video compares, In, M is positive integer.
The processing that video provided by Embodiment 2 of the present invention compares can be performed in the processing unit that the video of the present embodiment compares Method, realization principle is similar, and details are not described herein again.
The present embodiment by executing following mistake for first via vision signal to be compared and the second tunnel vision signal respectively Journey: present frame figure is determined according to the image feature value of each pixel of current frame image for each frame image of vision signal The fit characteristic value of each comparison pel of picture；Wherein, respectively compare pel be by a frame image be divided into predetermined number region and It obtains；The fit characteristic value for determining each comparison pel of the first via vision signal on each frame image, with the second road video Whether the fit characteristic value of each comparison pel of the signal on each frame image is identical, to find first via vision signal and second The video frame synchronization point of road vision signal, wherein video frame synchronization point characterizes first via vision signal and the second tunnel vision signal It is initially synchronous from the video frame synchronization point；For each frame image since video frame synchronization point, determine that the first via regards The fit characteristic value of each comparison pel of the frequency signal on current frame image is believed with the second tunnel vision signal with first via video Number current frame image corresponding to each comparison pel on image fit characteristic value between each difference, and according to each difference Determine whether first via vision signal and the second tunnel vision signal are consistent on each frame image.Wherein, it is wrapped in image feature value Y characteristic value, U characteristic value and V characteristic value are included.To provide a kind of new video comparison method, by two-path video signal into The comparison of row vision signal is analyzed, and whether extremely can more accurately monitor broadcast vision signal；Also, needle The division that region is carried out to each frame image of vision signal obtains at least one and compares pel, using the fitting for comparing pel Characteristic value analyzes the consistency of each frame image, and the video analysis method accurate and effective provided can be identified effectively Whether each frame image of vision signal has occurred exception out, and then effectively identifies whether vision signal is abnormal；And And it is mainly based upon what Y characteristic value was compared compared to existing video alignments, and since video is in encoding and decoding, biography The complexity of defeated equal links process, the comparison based on Y characteristic value, which often exists, compares wrong report, in this application to Y characteristic value, U Characteristic value and V characteristic value are analyzed, and vector characteristic value is obtained, then obtain a fit characteristic value, and then can be special based on YUV Value indicative carries out analysis comparison to video, can effectively reduce wrong report, improves and compares accuracy.
Those of ordinary skill in the art will appreciate that: realize that all or part of the steps of above-mentioned each method embodiment can lead to The relevant hardware of program instruction is crossed to complete.Program above-mentioned can be stored in a computer readable storage medium.The journey When being executed, execution includes the steps that above-mentioned each method embodiment to sequence；And storage medium above-mentioned include: ROM, RAM, magnetic disk or The various media that can store program code such as person's CD.
Finally, it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations；Although Present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that: it still may be used To modify the technical solutions described in the foregoing embodiments or equivalent replacement of some of the technical features； And these are modified or replaceed, technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution spirit and Range.

Claims (16)

1. the processing method that a kind of video compares characterized by comprising
For first via vision signal to be compared and the second tunnel vision signal, following procedure is executed respectively: for the video Each frame image of signal determines each comparison of current frame image according to the image feature value of each pixel of current frame image The fit characteristic value of pel；Wherein, each comparison pel is that a frame image is divided into obtained from the region of predetermined number；
The fit characteristic value for determining each comparison pel of the first via vision signal on each frame image, with second tunnel Whether the fit characteristic value of each comparison pel of the vision signal on each frame image is identical, to find the first via video letter Video frame synchronization point number with second tunnel vision signal, wherein the video frame synchronization point characterizes the first via video Signal and second tunnel vision signal from the video frame synchronization point be initially synchronous；
Each frame image since the video frame synchronization point determines for each comparison pel of each frame image The fit characteristic value of the current comparison pel of the first via vision signal and the current comparison chart of second tunnel vision signal Difference between the fit characteristic value of member；
Each frame image since the video frame synchronization point, for each comparison pel of each frame image, true When the fixed difference is less than or equal to difference threshold, the current comparison pel and described the of the first via vision signal is determined The current comparison pel of two tunnel vision signals is consistent；When determining that the difference is greater than difference threshold, described the is determined The current comparison pel of vision signal and the current comparison pel of second tunnel vision signal are inconsistent all the way, and The inconsistent block number of pel is compared in current frame image；
Each frame image since the video frame synchronization point is determining the comparison pel for each frame image When inconsistent block number is less than third block number threshold value, determine that the first via vision signal and second tunnel vision signal exist It is consistent on current frame image；When determining that the inconsistent block number of the comparison pel is more than or equal to third block number threshold value, Determine that the first via vision signal and second tunnel vision signal are inconsistent on current frame image；
Each frame image since the video frame synchronization point is determining the first via vision signal and second tunnel view It is abnormal to determine that video compares when being inconsistent on continuous Z frame image for frequency signal, and issues the abnormal information warning of comparison, wherein Z is positive integer；
After issuing and comparing abnormal information warning, determining that the first via vision signal and second tunnel vision signal exist When being consistent on continuous N frame image, it is normal to determine that video compares, wherein M is positive integer.
2. the method according to claim 1, wherein each frame image for the vision signal, root According to the image feature value of each pixel of current frame image, the fit characteristic value of each comparison pel of current frame image is determined, wrap It includes:
For each frame image, by the pixel of every X adjacent position of current frame image, it is determined as a foundation drawing Member, wherein X is positive integer；
For each basic pel of each frame image, according to the characteristics of image of each pixel in the basis pel Value, determines the foundation characteristic value of the basis pel；
For each frame image, by the basic pel per N number of adjacent position of current frame image, it is determined as a comparison chart Member, wherein N is positive integer；
For each comparison pel of each frame image, according to the foundation characteristic of each basic pel in the comparison pel Value, determines the vector characteristic value of the comparison pel；
For each comparison pel of each frame image, according to the vector characteristic value of the comparison pel, the comparison is determined The fit characteristic value of pel.
3. according to the method described in claim 2, it is characterized in that, including Y characteristic value, U feature in described image characteristic value Value and V characteristic value；It include Y characteristic mean, U characteristic mean and V characteristic mean in the foundation characteristic value；
It is special according to the image of each pixel in the basis pel for described each basic pel for each frame image Value indicative determines the foundation characteristic value of the basis pel, comprising:
For each basic pel of each frame image, the Y characteristic value of each pixel in the basis pel is determined Mean value, the mean value of the U characteristic value of each pixel, the mean value of the V characteristic value of each pixel, respectively the Y feature of the basis pel Mean value, U characteristic mean, V characteristic mean.
4. according to the method described in claim 3, it is characterized in that, including Y ' characteristic value, U ' spy in the vector characteristic value Value indicative and V ' characteristic value；
For described each comparison pel for each frame image, according to the basis of each basic pel in the comparison pel Characteristic value determines the vector characteristic value of the comparison pel, comprising:
For each comparison pel of each frame image, by the Y characteristic mean of each basic pel in the comparison pel Radian value conversion is carried out, obtains the Y ' characteristic value of the comparison pel, and the U feature of each basic pel in the comparison pel is equal Value carries out radian value conversion, obtains the U ' characteristic value of the comparison pel, and by the V feature of each basic pel in the comparison pel Mean value carries out radian value conversion, obtains the V ' characteristic value of the comparison pel.
5. according to the method described in claim 4, it is characterized in that, the fit characteristic value be P=a*Y '+b*U '+c*V ', In, a, b and c are weighting coefficient.
6. according to the method described in claim 3, it is characterized in that, in each frame image for the vision signal, According to the image feature value of each pixel of current frame image, determine each comparison pel of current frame image fit characteristic value it Before, further includes:
When the vision signal is SD vision signal, for each frame image of the vision signal, by current frame image Y, U, V SD characteristic value, be converted to R, G, B characteristic value, and according to preset compensation factor, R, G, B characteristic value is turned It is changed to Y, U, V high definition characteristic value；Wherein, Y, U, V high definition characteristic value is the image of each pixel of current frame image Characteristic value.
7. according to the method described in claim 6, it is characterized in that, the Y high definition characteristic value in described image characteristic value is 0.21* R+0.71*G+0.07*B, the U high definition characteristic value in described image characteristic value are (0.5*B-0.11*R-0.38*G) * 1.02+ 128, the V high definition characteristic value in described image characteristic value is (0.5*R-0.45*G-0.045*B) * 1.02+128.
8. method according to claim 1-7, which is characterized in that the determination first via vision signal exists The fit characteristic value of each comparison pel on each frame image, with each ratio of second tunnel vision signal on each frame image It is whether identical to the fit characteristic value of pel, to find the video of the first via vision signal Yu second tunnel vision signal Frame synchronization point, comprising:
The fit characteristic value for determining each comparison pel of the first via vision signal on each frame image, with second tunnel Whether the fit characteristic value of each comparison pel of the vision signal on each frame image is identical, with the determination first via video letter Number with second tunnel vision signal on each frame image whether respectively compare pel identical, and determine the first via video The signal block number different from comparison pel of second tunnel vision signal on each frame image；
Determining that the first via vision signal compares in each frame image of continuous P frame image with second tunnel vision signal The block number different to pel when less than the first block number threshold value, determines that the frame after P frame is the video frame synchronization point；Alternatively, Determining the first via vision signal and second tunnel vision signal comparison chart in each frame image of continuous Q frame image Member different block number when less than the second block number threshold value, determines that the frame after Q frame is the video frame synchronization point；
Wherein, P, Q are positive integer, and P is greater than Q, and the first block number threshold value is less than the second block number threshold value.
9. the processing unit that a kind of video compares characterized by comprising
Determining module, for executing following procedure respectively for first via vision signal to be compared and the second tunnel vision signal: Present frame is determined according to the image feature value of each pixel of current frame image for each frame image of the vision signal The fit characteristic value of each comparison pel of image；Wherein, each comparison pel is that a frame image is divided into predetermined number Obtained from region；
Analysis module, for determining the fit characteristic of each comparison pel of the first via vision signal on each frame image Value, it is whether identical as the fit characteristic value that respectively compares pel of second tunnel vision signal on each frame image, to find The video frame synchronization point of the first via vision signal and second tunnel vision signal, wherein the video frame synchronization point table Levy the first via vision signal and second tunnel vision signal from the video frame synchronization point be initially synchronous；
First compares submodule, for each frame image since the video frame synchronization point, for the every of each frame image One compare pel for, determine the current comparison pel of the first via vision signal fit characteristic value and second tunnel Difference between the fit characteristic value of the current comparison pel of vision signal；
Second compares submodule, for each frame image since the video frame synchronization point, for the every of each frame image One compares for pel, when determining that the difference is less than or equal to difference threshold, determines working as the first via vision signal Preceding comparison pel and the current comparison pel of second tunnel vision signal are consistent；Determining the difference greater than difference When threshold value, the current comparison pel of the first via vision signal and the current comparison of second tunnel vision signal are determined Pel is inconsistent, and the inconsistent block number of pel is compared in current frame image；
Third compares submodule, for each frame image since the video frame synchronization point, for each frame image, When determining that the inconsistent block number of the comparison pel is less than third block number threshold value, the first via vision signal and institute are determined The second tunnel vision signal is stated to be consistent on current frame image；Determining that the inconsistent block number of the comparison pel is more than or equal to When third block number threshold value, determine that the first via vision signal and second tunnel vision signal are not on current frame image It is consistent；
First confirmation submodule is determining the first via view for each frame image since the video frame synchronization point It is abnormal to determine that video compares when being inconsistent on continuous Z frame image for frequency signal and second tunnel vision signal, and issue ratio To abnormal information warning, wherein Z is positive integer；
Second confirmation submodule is used for after the first confirmation submodule issues and compares abnormal information warning, determining First via vision signal and second tunnel vision signal are stated when being consistent on continuous N frame image, it is normal to determine that video compares, Wherein, M is positive integer.
10. device according to claim 9, which is characterized in that the determining module, comprising:
First determines submodule, for being directed to for each frame image, by the pixel of every X adjacent position of current frame image Point is determined as a basic pel, wherein X is positive integer；
Second determines submodule, for each basic pel for each frame image, according in the basis pel The image feature value of each pixel determines the foundation characteristic value of the basis pel；
Third determines submodule, for being directed to for each frame image, by the foundation drawing per N number of adjacent position of current frame image Member is determined as a comparison pel, wherein N is positive integer；
4th determines submodule, for comparing for pel for each of each frame image, according in the comparison pel The foundation characteristic value of each basis pel, determines the vector characteristic value of the comparison pel；
5th determines submodule, for comparing for pel for each of each frame image, according to the arrow of the comparison pel Measure feature value determines the fit characteristic value of the comparison pel.
11. device according to claim 10, which is characterized in that include Y characteristic value, U spy in described image characteristic value Value indicative and V characteristic value；It include Y characteristic mean, U characteristic mean and V characteristic mean in the foundation characteristic value；
Described second determines submodule, is specifically used for:
For each basic pel of each frame image, the Y characteristic value of each pixel in the basis pel is determined Mean value, the mean value of the U characteristic value of each pixel, the mean value of the V characteristic value of each pixel, respectively the Y feature of the basis pel Mean value, U characteristic mean, V characteristic mean.
12. device according to claim 11, which is characterized in that include Y ' characteristic value, U ' in the vector characteristic value Characteristic value and V ' characteristic value；
Described 4th determines submodule, is specifically used for:
For each comparison pel of each frame image, by the Y characteristic mean of each basic pel in the comparison pel Radian value conversion is carried out, obtains the Y ' characteristic value of the comparison pel, and the U feature of each basic pel in the comparison pel is equal Value carries out radian value conversion, obtains the U ' characteristic value of the comparison pel, and by the V feature of each basic pel in the comparison pel Mean value carries out radian value conversion, obtains the V ' characteristic value of the comparison pel.
13. device according to claim 12, which is characterized in that the fit characteristic value is P=a*Y '+b*U '+c*V ', Wherein, a, b and c are weighting coefficient.
14. device according to claim 11, which is characterized in that described device, further includes:
Conversion module, for being directed to each frame image of the vision signal in the determining module, according to current frame image The image feature value of each pixel, before the fit characteristic value of each comparison pel for determining current frame image, for be compared First via vision signal and the second tunnel vision signal, execute following procedure respectively: being SD vision signal in the vision signal When, Y, U, V SD characteristic value of current frame image are converted into R, G, B feature for each frame image of the vision signal Value, and according to preset compensation factor, R, G, B characteristic value is converted into Y, U, V high definition characteristic value；Wherein, described Y, U, V High definition characteristic value is the image feature value of each pixel of current frame image.
15. device according to claim 14, which is characterized in that the Y high definition characteristic value in described image characteristic value is 0.21*R+0.71*G+0.07*B, the U high definition characteristic value in described image characteristic value are (0.5*B-0.11*R-0.38*G) * 1.02+128, the V high definition characteristic value in described image characteristic value are (0.5*R-0.45*G-0.045*B) * 1.02+128.
16. according to the described in any item devices of claim 9-15, which is characterized in that the analysis module is specifically used for:
The fit characteristic value for determining each comparison pel of the first via vision signal on each frame image, with second tunnel Whether the fit characteristic value of each comparison pel of the vision signal on each frame image is identical, with the determination first via video letter Number with second tunnel vision signal on each frame image whether respectively compare pel identical, and determine the first via video The signal block number different from comparison pel of second tunnel vision signal on each frame image；
Determining that the first via vision signal compares in each frame image of continuous P frame image with second tunnel vision signal The block number different to pel when less than the first block number threshold value, determines that the frame after P frame is the video frame synchronization point；Alternatively, Determining the first via vision signal and second tunnel vision signal comparison chart in each frame image of continuous Q frame image Member different block number when less than the second block number threshold value, determines that the frame after Q frame is the video frame synchronization point；
Wherein, P, Q are positive integer, and P is greater than Q, and the first block number threshold value is less than the second block number threshold value.
CN201710502377.0A 2017-06-27 2017-06-27 The treating method and apparatus that video compares Active CN107277500B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710502377.0A CN107277500B (en) 2017-06-27 2017-06-27 The treating method and apparatus that video compares

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710502377.0A CN107277500B (en) 2017-06-27 2017-06-27 The treating method and apparatus that video compares

Publications (2)

Publication Number Publication Date
CN107277500A CN107277500A (en) 2017-10-20
CN107277500B true CN107277500B (en) 2019-03-19

Family

ID=60069851

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710502377.0A Active CN107277500B (en) 2017-06-27 2017-06-27 The treating method and apparatus that video compares

Country Status (1)

Country Link
CN (1) CN107277500B (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110020588A (en) * 2019-01-28 2019-07-16 深圳市睿智物联科技有限公司 A kind of building screen lamps and lanterns monitoring method and system based on video identification
WO2022150347A1 (en) * 2021-01-05 2022-07-14 Google Llc Subsurface display interfaces and associated systems and methods
CN112954408B (en) * 2021-01-25 2023-05-12 中央广播电视总台 Abnormality monitoring method and corresponding device for ultra-high definition video signal

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1107177A2 (en) * 1999-12-09 2001-06-13 ATI International SRL Method and apparatus for processing portions of primitives that are being rendered
CN103067778A (en) * 2013-01-06 2013-04-24 北京华兴宏视技术发展有限公司 Data monitoring system and data monitoring method
CN104539975A (en) * 2014-12-03 2015-04-22 上海交通大学 Double-compressed video tampering detection method based on blocking effect
CN106507191A (en) * 2016-11-15 2017-03-15 北京鹏润鸿途科技股份有限公司 Synchronous relay content monitoring method and device

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1107177A2 (en) * 1999-12-09 2001-06-13 ATI International SRL Method and apparatus for processing portions of primitives that are being rendered
CN103067778A (en) * 2013-01-06 2013-04-24 北京华兴宏视技术发展有限公司 Data monitoring system and data monitoring method
CN104539975A (en) * 2014-12-03 2015-04-22 上海交通大学 Double-compressed video tampering detection method based on blocking effect
CN106507191A (en) * 2016-11-15 2017-03-15 北京鹏润鸿途科技股份有限公司 Synchronous relay content monitoring method and device

Also Published As

Publication number Publication date
CN107277500A (en) 2017-10-20

Similar Documents

Publication Publication Date Title
CN107277500B (en) The treating method and apparatus that video compares
CN106170087B (en) Anomalous pixel detection
US20070217765A1 (en) Method and its application for video recorder and player
US10769849B2 (en) Use of temporal motion vectors for 3D reconstruction
CN103067740B (en) Video monitoring equipment intelligent fault detection method and detection system thereof
US20110135158A1 (en) Image processing device, image processing method and program
CN102395043A (en) Video quality diagnosing method
EP4090036A1 (en) Privacy shielding processing method and apparatus, electronic device, and monitoring system
CN104702897B (en) The quick history playback and the method for event latch that Internet video is made video recording
CN111028222B (en) Video detection method and device, computer storage medium and related equipment
CN111416960B (en) Video monitoring system based on cloud service
US8600157B2 (en) Method, system and computer program product for object color correction
CN105430382A (en) Method and device for detecting black edge of video image
CN102957940A (en) Improvement-based four-way operator video noise detection method
JP5710230B2 (en) Monitoring system and monitoring method
US8502922B2 (en) Multimedia device and play mode determination method of the same
CN107734278A (en) A kind of video playback method and relevant apparatus
CN108737816A (en) A kind of black field detecting method of IPTV
CN106998464B (en) Detect the method and device of thorn-like noise in video image
CN111399476B (en) Real-time detection method for monitoring out-of-control holder based on image alignment
US20160012607A1 (en) Method and apparatus to detect artificial edges in images
CN102196259B (en) Moving object detection system and method suitable for compression domain
CN107197326B (en) Video frequency identifying method and device
CN117041666A (en) Abnormality detection method and device for digital video, electronic equipment and storage medium
CN106331649B (en) Video data copy, joining method and system

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant