CN103413439A - Method for sorting passenger vehicles and goods vehicles based on videos - Google Patents
Method for sorting passenger vehicles and goods vehicles based on videos Download PDFInfo
- Publication number
- CN103413439A CN103413439A CN2013103264558A CN201310326455A CN103413439A CN 103413439 A CN103413439 A CN 103413439A CN 2013103264558 A CN2013103264558 A CN 2013103264558A CN 201310326455 A CN201310326455 A CN 201310326455A CN 103413439 A CN103413439 A CN 103413439A
- Authority
- CN
- China
- Prior art keywords
- video
- vehicle
- image
- camera
- led lamp
- 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.)
- Granted
Links
Images
Landscapes
- Image Processing (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention provides a method for sorting passenger vehicles and goods vehicles based on videos. The method uses LED lamp bands as an auxiliary, and is based on the theory of specular reflection of car window glass, according to outstanding features that the window number of a goods vehicle and the window number of a passenger vehicle are different, through the algorithms of vehicle tracking, binarization processing, banded target detecting and the like, the method obtains the reflected images of the target LED lamp bands in a video sequence, achieves detection and identification of whether car window glass exists, judges whether enough car window glass exists, therefore real-time and reliable division on the types of large and medium passenger vehicles and the goods vehicles is carried out, and due to the fact that the technology carries out detection through a non-contact mode, the failure rate is low. Special passageways do not need to be arranged for the passenger vehicles and the goods vehicles, the utilizing rate of limited space is improved greatly, traffic will not be blocked when mounting and maintenance are carried out, and the method for sorting the passenger vehicles and the goods vehicles based on the videos has wide application prospect in highway tolling systems.
Description
Technical field
The invention belongs to video and detect and technical field of information processing, be specifically related to a kind of sorting technique of lorry and passenger vehicle based on video.
Background technology
In the highway tolling system of China, to the expenses standard system of lorry and passenger vehicle, be different, wherein, its pay load of charge Main Basis of lorry, the charge of passenger vehicle the Main Basis vehicle kind or appraise and decide passengers quantity.Therefore, in Fare Collection System, must at first distinguish lorry and passenger vehicle.
At present, the Expressway Toll Methods of China mainly contains semi automatic toll and ETC(non-parking charge) two kinds.In the semi automatic toll mode, main by artificial lorry and the passenger vehicle distinguished; And in ETC, in order to distinguish lorry and passenger vehicle, the general employing arranges special passenger vehicle and lorry passage, to realize the purpose of lorry and passenger vehicle difference charge.
No matter be semi automatic toll or ETC, if can realize the automatic identification of lorry and passenger vehicle, will greatly improve charge efficiency and management level.For example, the lane in which the drivers should pay fees that some flow is little can be realized unattended self-service charge, lorry and passenger vehicle also can arrange separately special charge passage, so not only use manpower and material resources sparingly, take full advantage of path resource, and be convenient to unified management, make Fare Collection System more improve with efficient.As can be seen here, the automatic identification technology of research passenger vehicle and lorry, be of great practical significance for the automatization level that improves highway tolling system.
Current vehicle classification technology mainly contains the pressure transducer sorting technique, laser detection sorting technique, infrared detection sorting technique, the electromagnetic induction coil sorting technique, wireless telecommunications sorting technique etc., the system of these methods and specific classification standard are closely related, thereby system is portable poor; The system of adopts pressure sensor or electromagnetic induction coil also needs road surface pavement again, inconvenience is installed, lacks dirigibility.These systems also have a common shortcoming, are exactly because equipment work under bad environment, serviceable life are limited, need all vehicles that impulse senders are installed based on the vehicle identification system of wireless communication technique, invest very large.
Summary of the invention
Present situation for highway tolling system, the object of the invention is to, and proposes a kind of method of passenger vehicle based on video and lorry classification, and the method can realize in real time the big-and-middle-sized lorry in range of video and passenger vehicle, classification reliably.
In order to realize above-mentioned technical assignment, the technical solution used in the present invention is:
A kind of passenger vehicle based on video and lorry sorting technique, the method utilizes computing machine the sequence of video images of camera acquisition to be processed to the classification that realizes passenger vehicle and lorry, on described computing machine, be connected with the infrared vehicle separation vessel, the infrared vehicle separation vessel is arranged on freeway toll station feeder connection place, a side in the close charge station of infrared vehicle separation vessel direction is equipped with LED lamp band, vertically arrange, in LED lamp band bottom, the first video camera is installed, one side at LED lamp band middle part is equipped with the second video camera, the camera lens of the first video camera and the second video camera all is parallel to road surface, and vertical with road direction, described the first video camera is connected with computing machine by an image pick-up card respectively with the second video camera, should comprise the following steps based on passenger vehicle and the lorry sorting technique of video:
Step 1, the first video camera and the second video camera catch video information within the vision, computing machine carries out image acquisition to the video information of the first video camera and the second camera acquisition respectively by image pick-up card, obtains the sequence of video images of the first video camera and the second camera acquisition;
Whether step 2, have the vehicle process on infrared vehicle separation vessel monitoring road, if the headstock process of vehicle detected, this information passed to computing machine, and note video frame number is N, and video adds up to M, and the initial value of M and N is 0; Execution step three;
Step 3, the computing machine utilization is processed the sequence of video images of the first camera acquisition based on the target tracking algorism of feature angle point:
(1) the first camera acquisition to sequence of video images in, selecting video image sequence current frame image, adopt the Moravec algorithm to extract Corner, and using centered by this angle point and choose an image block as template;
(2) adopt and follow the tracks of based on Block Matching Algorithm, in the next frame image of present frame, it is poor that the pixel value of piece that will be onesize with above-mentioned template correspondence position and the pixel value of template are done, if the number that margin of image element is not 0 is greater than 5, perform step four, continue to carry out otherwise return to step 3;
Step 4, computing machine to the second camera acquisition to sequence of video images process:
(1) demarcation of target area
The second camera acquisition to sequence of video images in, the computer selecting current frame image, be divided into three zones that size is identical by current frame image with vertical direction, the zone in the middle of choosing is as processing region;
(2) adopt the background subtraction point-score to carry out cutting apart of target to processing region
Each pixel to the processing region chosen adopts global threshold binarization method to process, and the zone after then processing is divided into the piece that a plurality of sizes are identical, then each piece is carried out to block-based binary conversion treatment;
(3) connected component labeling
Adopt eight neighborhood labeling algorithms to carry out connected component labeling to the image block after binary conversion treatment, and carry out the removal of connected domain and fill to merge obtaining Contiguous graphics; The closed operation of again Contiguous graphics being carried out in morphologic filtering is processed, and obtains destination object;
(4) detection of Band object
The geometric characteristic of evaluating objects object, the depth-width ratio value of calculating destination object, according to ratio, if the depth-width ratio value of destination object is greater than 25, determine that this target is Band object;
(5) tracking of Band object statistics
If destination object is not Band object, the video sum adds 1; If destination object is Band object, video frame number N adds 1, and the video sum adds 1; If the infrared vehicle separation vessel detects vehicle tail by the infrared vehicle separation vessel, perform step five, otherwise return to step 3;
Further, the length of described LED lamp band is 1.5～2m, and width is 0.015m, and the bottom of LED lamp band is 1.2～2m apart from the height on ground, and the light intensity of LED lamp band is 2500～3000cd.
Further, after in step 3, adopting the Moravec algorithm to extract Corner, using centered by this angle point choose a size as the image block of 25*25 pixel as template.
The present invention is based on lorry and the passenger vehicle sorting technique of video, is a kind of novel video detection technology, can accurately to the large and medium bus in range of video and lorry, classify, and is not subjected to environmental restraint, and real-time is good, and be easy to realize, accuracy is higher.Compared with prior art, this technology adopts contactless mode to detect, failure rate is low; Passenger vehicle and car needn't arrange special passage, have greatly improved the utilization factor of the finite space; When installing and keeping in repair, can not block traffic, have broad application prospects in highway tolling system.
The accompanying drawing explanation
Fig. 1 is the artwork of equipment in the present invention;
Fig. 2 is overall flow figure of the present invention;
Fig. 3 is the mirror-reflection schematic diagram;
Fig. 4 utilizes the Moravec algorithm to extract the schematic diagram of an angle point on image;
Fig. 5 (a) is the image of LED lamp band of the Bus window reflection of the second camera acquisition;
Fig. 5 (b) is the image of the truck body reflection LED lamp band of the second camera acquisition;
Fig. 6 (c) is the result that Fig. 5 (a) adopts the global threshold binarization method to process;
Fig. 6 (d) is the result that Fig. 5 (b) adopts the global threshold binarization method to process;
Fig. 7 (e) is the image after the block-based binaryzation of Fig. 5 (a), connected component labeling;
Fig. 7 (f) is the image after the block-based binaryzation of Fig. 5 (b), connected component labeling;
Fig. 8 is the schematic diagram of simulation tracing statistics Band object;
The present invention is described in further detail below in conjunction with drawings and Examples.
Embodiment
The present embodiment provides a kind of method of passenger vehicle based on video and lorry classification, mainly by detecting car body, whether has abundant vehicle window, distinguishes passenger vehicle and lorry with this.Specific implementation is to take LED lamp band as auxiliary, utilize the principle of the mirror-reflection of glass for vehicle window, according to lorry this distinguishing features different from the Bus window number, by analyzing the reflected image of target LED lamp band in video sequence, with realization, to whether having the detection identification of glass for vehicle window, and judge whether abundant glass for vehicle window, with this, passenger vehicle and lorry kind are carried out in real time, reliable division, as shown in Figure 1 to Figure 3; Specifically follow these steps to carry out:
The selection of equipment:
Before carrying out the equipment installation, first to the length of existing passenger vehicle and lorry with highly carry out statistical study so that suitable selected equipment with carry out the installation of equipment.In Table 1 and table 2:
The existing lorry specification of table 1
The existing passenger vehicle specification of table 2
Annotate: " the vehicle window overall length " in table 1 and table 2 refers to that all vehicle windows of a side of vehicle follow the length sum of car direction (length that does not comprise vehicle window midfeather part)
By large quantitative statistics, learnt, the height of existing passenger vehicle and lorry is between 1.8 meters to 3.7 meters, the height of glass for vehicle window is between 0.5 meter to 1.5 meters, therefore the length of LED lamp band is between 1.5 meters to 2 meters, in glass for vehicle window, just can clearly reflex to like this image of lamp band, in order to reduce the impact of surrounding enviroment on system, lamp band brightness range of choice is 2500～3000 candelas.
This system coordinates the existing facility of Current Highway charge station to use, at the freeway toll station feeder connection, the echelette vehicle separator is installed, a side in the close charge station of infrared vehicle separation vessel direction is equipped with described LED lamp band, LED lamp band bottom is high to 2 meters apart from 1.2 meters, road surface, vertically install, the first video camera is arranged on to LED lamp band bottom, camera lens is parallel to level road, vertical with road direction, catch the car body through vehicle; Road direction, be the direction of vehicle movement; A side in lamp band centre position is installed the second video camera, and camera lens is parallel to horizontal direction, vertical with road direction, the vehicle window image of process vehicle on capture channel; Then by image pick-up card, gather image, the image collected is passed to the master-control room computing machine and carry out follow-up processing, the first video camera and the second video camera are used in conjunction with, as shown in Figure 2.
The main vehicle body image of the image of the first camera acquisition, be used to judging whether vehicle, by the infrared vehicle separation vessel time, has situation about stopping, and the second video camera catches, it is the frame number of image of the LED lamp band of vehicle window reflection, and the totalframes that gathers during by the second video camera of vehicle integral body, the ratio of the totalframes while passing through with vehicle is whole by the frame number that LED lamp band reflected image is arranged, can judge that what vehicle the vehicle passed through is, therefore, when vehicle when by vehicle separator, stopping, if computing machine continues to carry out to the image processing process of the second camera acquisition, there is the frame number of LED lamp band reflected image not increase, but the video totalframes that the second camera acquisition arrives is increasing always, can cause the inaccurate of result.Therefore to start the prerequisite that the image of the second camera acquisition processes be that vehicle is kept in motion to computing machine, and this computer-chronograph just can be added up the totalframes that the frame number that comprises LED lamp band reflected image and vehicle integral body are passed through.
Step 1, the first video camera and the second video camera catch video information within the vision, computing machine carries out image acquisition to the video information of the first video camera and the second camera acquisition respectively by image pick-up card, obtains the sequence of video images of the first video camera and the second camera acquisition;
Whether step 2, have the vehicle process on infrared vehicle separation vessel monitoring road, if the headstock process of vehicle detected, this information passed to computing machine, computing machine start to camera acquisition to image information process.Note video frame number is N, and video adds up to M, and the initial value of M and N is 0; Execution step three;
Step 3, the computing machine utilization is processed the sequence of video images of the first camera acquisition based on the target tracking algorism of feature angle point, realizes the tracking of vehicle, is kept in motion to guarantee vehicle, if vehicle is not in motion state, do not perform step four so;
(1) adopt classical Moravec algorithm to extract angle point, the Moravec algorithm utilizes the variance of gray scale to extract angle point, this algorithm is by definition " interest value ", carry out on this basis closed operation processing and non-maximal value and suppress to calculate angle point, the angle point obtained as shown in Fig. 4 (image centre circle in point), being implemented as follows of Morave algorithm:
With a certain pixel (x on image, y) centered by, set up the window that a size is n*n (such as the window of 5*5), as shown in Figure 4, when window transversely, vertically and two cornerwise directions while moving, calculate respectively interest value on four direction (quadratic sum of neighbor gray scale difference) g
1, g
2, g
3, g
4, expression formula is as follows:
In formula, k=INT (n/2) (n/2 is rounded, and n is window width); g
1Mean horizontal interest value, g
2Mean vertical interest value, g
3And g
4Mean two diagonal interest value, i is mobile number of pixels, and k is mobile ultimate range.From g
1, g
2, g
3, g
4Middle selected value minimum value is as the metric of this interest value.First the interest value of each point on image is selected, then in all these interest value, selected to be greater than the point of a certain threshold value as angle point.
The direction gradient difference of larger this pixel of explanation of interest value is larger, and the quality of angle point is better.Interest value is greater than to threshold value T(T desirable 120～130) point screen as angle point, using centered by this point choose a size as the piece of 25*25 as template.
(2) adopt and follow the tracks of based on Block Matching Algorithm, the method of piece coupling is as follows: be located in current frame image, with the Moravec algorithm, an angle point P (x detected, y), in the next frame image of present frame, the pixel value of piece that will be onesize with above-mentioned template correspondence position and the pixel value of template are poor, margin of image element is not that 0 number p is greater than 5, if (two two field pictures there are differences, in can selected digital image, a certain size piece compares the comparison of pixel value, through test of many times, find, the piece of selected 25*25 pixel compares, the number that pixel changes is less than 5～10, think that this two two field picture not there are differences, here select 5) perform step four, otherwise continue execution step three.
Employing is followed the tracks of based on Block Matching Algorithm, so when vehicle stops in the process by the infrared vehicle separation vessel, angle point on angle point on present image and next frame image does not change, and computing machine does not start the work for the treatment of of the image that the second camera acquisition arrives; If vehicle does not remain static, the corner location information of present frame and next frame is different so, reaction is margin of image element position 0 on margin of image element number p is greater than 5, now namely judges state of motion of vehicle, computing machine start to the second camera acquisition to image process.
Step 4, computing machine to the second camera acquisition to sequence of video images process:
Carry out the Band object statistics of LED lamp band reflected image, be implemented as follows:
(1) demarcation of target area.
In the sequence of video images of the second camera acquisition, the spotting surveyed area, the speed of processing in order to improve image, current frame image is divided into to three zones that size is identical with vertical direction, choosing the zone of whole two field picture middle 1/3rd processes, as shown in Figure 5, (a) for the Bus window collected, reflect the image of LED lamp band, LED lamp band is apparent in view; (b) be the LED lamp band image of truck body reflection, the vehicle body reflection is not obvious;
(2) adopt the background subtraction point-score to carry out cutting apart of target
Utilize the background image (when vehicles failed arrives, the image of the current location of the second camera acquisition) B (x, y) and present frame target area image f (x that have obtained, y) poor, then with the threshold value of setting, make comparisons, thereby obtain target information D (x, y), as shown in the formula:
T wherein
2(between 150 to 180) threshold value for setting, in image, D (x, y)=255(0 represents black, 255 represents white) the pixel representative be object pixel, i.e. the result of each on image point employing global threshold binarization method processing, as shown in Figure 6.On this basis image is divided into to the piece of a plurality of number sizes; The pixel size of this video sequence is that 720*288 can be divided into the big or small piece of 8*6 that is, add up the number of white pixel in each piece, if be greater than 1/3rd of number of pixels sum in this piece, just all make all pixel values in this piece into 255, completed block-based binary conversion treatment;
(3) mark of connected domain
Connectedness between image pixel is to determine the key concept of object boundary information and area information in image.Whether two pixels connect needs is confirmed whether they contact in some sense.Adopt eight neighborhood labeling algorithms to carry out connected component labeling to the image block after binary conversion treatment, eight neighborhood labeling algorithms are specific as follows:
If the some set { (x+p, y+q) corresponding with pixel (x, y); , p and q are a pair of significant integer, are referred to as the neighborhood of pixel (x, y).At first define 8 neighborhoods of pixel (x, y), as follows:
F8(x,y)＝{f(x-1,y-1),f(x-1,y),f(x-1,y+1),f(x,y-1),
f(x,y+1),f(x+1,y+1),f(x+1,y),f(x+1,y+1)}
If pixel f (x, y) and f (x+p, y+q) are arranged in same width image-region, if f (x, y) and f (x+p, y+q) exist 8 vertex neighborhoods to be communicated with, claim f (x, y) and f (x+p, y+q) to be communicated with.The object of connected domain research is not pixel but image block, if two image blocks meet 8 Neighbor Conditions and just think that the two is connected, be labeled as same target, again the target after mark is removed, fills, merged and obtain Contiguous graphics, then the closed operation of Contiguous graphics being carried out in morphologic filtering is processed, and obtains destination object.
(4) Band object detects
Analyze after treated and obtain destination object, calculate the depth-width ratio value of destination object, because the width of LED lamp band is 15mm, the height of the lamp band becomes in glass for vehicle window image is more than or equal to 1/2nd of vehicle window height, if 1/2nd and lamp bandwidth ratio range of vehicle window height are between 25 to 40, so the depth-width ratio value of object block is greater than at 25 o'clock, think that this target is Band object, be the result of LED lamp band after the reflected image that glass for vehicle window becomes is processed, as shown in (e) in Fig. 7; (f) due to the result that is the processing of vehicle body reflection LED lamp band image, so not Band object after the processing of LED lamp band reflectogram.
(5) tracking of Band object statistics
If destination object is not Band object, the video sum adds 1; If destination object is Band object, video frame number N adds 1, and the video sum adds 1; If now the infrared vehicle separation vessel detects vehicle tail and passes through, perform step so five, otherwise return to step 3; The video counts N here namely the second camera acquisition to vehicle on the frame number of vehicle window, the totalframes when video adds up to vehicle and crosses from the beginning to the end.
The M and the N value ratio calculated k that utilize step 4 to obtain, k=N/M.According to the Vehicle length of statistics in table 1 and table 2, to analyze, Bus window length and vehicle commander's ratio is greater than 90%, and lorry vehicle window length and vehicle commander's ratio is less than 20%, therefore when k was greater than 50%, this car was passenger vehicle, otherwise is lorry.As everyone knows, the glass for vehicle window number of passenger vehicle and lorry is different, and passenger vehicle has glass to exist from the headstock to the tailstock, and lorry only has headstock that glass is arranged, and according to the glass for vehicle window of car body, identifies also more directly perceived, easy.Therefore, can utilize glass for vehicle window to carry out the accurate classification of passenger vehicle and lorry.
Fig. 8 is simulation tracing statistics Band object process, and LED lamp band is static in whole process, and car body moves, and supposes that now car body is static, so just is equivalent to LED lamp band in motion, and vehicle window is scanned.When a car by the time, the totalframes while add up total number of image frames with appearance that glass for vehicle window is arranged, so just can calculate the number percent that the glass for vehicle window total length accounts for whole length over ends of body, finally carries out the classification of passenger vehicle and lorry.
It is below the specific embodiment that the inventor provides.
Embodiment:
At the freeway toll station feeder connection, model being installed is MYL infrared vehicle separation vessel, (model is: soft lamp band 220V60*5050 a length is installed is 2 meters, width in this position and be the LED lamp band of 15 millimeters, white,) lamp band bottom is high apart from 1.2 meters, road surface, the first video camera is arranged on to LED lamp band bottom, the specification that one side in LED lamp band centre position arranges second video camera the first video camera and the second video camera is that model is: gunlock, sharpness: 460 lines, lens focus: 8MM, monitoring camera lens: standard, photosensitive area: 1/3 cun.
In video sequence, when having a lorry to pass through, the N value of statistics is that 15, M value is 122 known, and calculating gained k value is 12%, is lorry thereby can judge it, as shown in Fig. 5 (a), conforms to actual.When having a passenger vehicle to pass through, the N value of statistics is that 63, M value is 57, and calculating gained k value is 90%, is passenger vehicle thereby can judge it, as shown in Fig. 5 (b), conforms to actual.
Claims (3)
1. the passenger vehicle based on video and lorry sorting technique, the method utilizes computing machine the sequence of video images of camera acquisition to be processed to the classification that realizes passenger vehicle and lorry, it is characterized in that, on described computing machine, be connected with the infrared vehicle separation vessel, the infrared vehicle separation vessel is arranged on freeway toll station feeder connection place, a side in the close charge station of infrared vehicle separation vessel direction is equipped with LED lamp band, vertically arrange, in LED lamp band bottom, the first video camera is installed, one side at LED lamp band middle part is equipped with the second video camera, the camera lens of the first video camera and the second video camera all is parallel to road surface, and vertical with road direction, described the first video camera is connected with computing machine by an image pick-up card respectively with the second video camera, should comprise the following steps based on passenger vehicle and the lorry sorting technique of video:
Step 1, the first video camera and the second video camera catch video information within the vision, computing machine carries out image acquisition to the video information of the first video camera and the second camera acquisition respectively by image pick-up card, obtains the sequence of video images of the first video camera and the second camera acquisition;
Whether step 2, have the vehicle process on infrared vehicle separation vessel monitoring road, if the headstock process of vehicle detected, this information passed to computing machine, and note video frame number is N, and video adds up to M, and the initial value of M and N is 0; Execution step three;
Step 3, the computing machine utilization is processed the sequence of video images of the first camera acquisition based on the target tracking algorism of feature angle point:
(1) the first camera acquisition to sequence of video images in, selecting video image sequence current frame image, adopt the Moravec algorithm to extract Corner, and using centered by this angle point and choose an image block as template;
(2) adopt and follow the tracks of based on Block Matching Algorithm, in the next frame image of present frame, it is poor that the pixel value of piece that will be onesize with above-mentioned template correspondence position and the pixel value of template are done, if the number that margin of image element is not 0 is greater than 5, perform step four, continue to carry out otherwise return to step 3;
Step 4, computing machine to the second camera acquisition to sequence of video images process:
(1) demarcation of target area
The second camera acquisition to sequence of video images in, the computer selecting current frame image, be divided into three zones that size is identical by current frame image with vertical direction, the zone in the middle of choosing is as processing region;
(2) adopt the background subtraction point-score to carry out cutting apart of target to processing region
Each pixel to the processing region chosen adopts global threshold binarization method to process, and the zone after then processing is divided into the piece that a plurality of sizes are identical, then each piece is carried out to block-based binary conversion treatment;
(3) connected component labeling
Adopt eight neighborhood labeling algorithms to carry out connected component labeling to the image block after binary conversion treatment, and carry out the removal of connected domain and fill to merge obtaining Contiguous graphics; The closed operation of again Contiguous graphics being carried out in morphologic filtering is processed, and obtains destination object;
(4) Band object detects
The geometric characteristic of evaluating objects object, the depth-width ratio value of calculating destination object, according to ratio, if the depth-width ratio value of destination object is greater than 25, determine that this target is Band object;
(5) tracking of Band object statistics
If destination object is not Band object, the video sum adds 1; If destination object is Band object, video frame number N adds 1, and the video sum adds 1; If the infrared vehicle separation vessel detects vehicle tail by the infrared vehicle separation vessel, perform step five, otherwise return to step 3;
Step 5, computing machine stop the processing of the sequence of video images that the first video camera and the second camera acquisition are arrived, now ratio calculated k=N/M; If k>50%, judge that the vehicle of firm process is passenger vehicle, otherwise be lorry.
2. the passenger vehicle based on video as claimed in claim 1 and lorry sorting technique, it is characterized in that, the length of described LED lamp band is 1.5～2m, and width is 0.015m, the bottom of LED lamp band is 1.2～2m apart from the height on ground, and the light intensity of LED lamp band is 2500～3000cd.
3. the passenger vehicle based on video as claimed in claim 1 and lorry sorting technique, is characterized in that, after in step 3, adopting the Moravec algorithm to extract Corner, using centered by this angle point choose a size as the image block of 25*25 pixel as template.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310326455.8A CN103413439B (en) | 2013-07-30 | 2013-07-30 | A kind of passenger vehicle based on video and lorry sorting technique |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310326455.8A CN103413439B (en) | 2013-07-30 | 2013-07-30 | A kind of passenger vehicle based on video and lorry sorting technique |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103413439A true CN103413439A (en) | 2013-11-27 |
| CN103413439B CN103413439B (en) | 2016-01-20 |
Family
ID=49606444
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310326455.8A Expired - Fee Related CN103413439B (en) | 2013-07-30 | 2013-07-30 | A kind of passenger vehicle based on video and lorry sorting technique |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103413439B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104574986A (en) * | 2014-12-22 | 2015-04-29 | 贵州中南交通科技有限公司 | Vehicle model identifier with radiating device |
| CN105702050A (en) * | 2016-04-22 | 2016-06-22 | 安徽皖通科技股份有限公司 | Highway over-limit and overload management control method |
| CN108171982A (en) * | 2018-02-06 | 2018-06-15 | 邓莉 | Crossing vehicular traffic vehicle monitor supervision platform |
| CN108417042A (en) * | 2017-09-18 | 2018-08-17 | 西安金路交通工程科技发展有限责任公司 | Car based on vehicle image and lorry sorting technique |
| CN111145365A (en) * | 2019-12-17 | 2020-05-12 | 北京明略软件***有限公司 | Method, device, computer storage medium and terminal for realizing classification processing |
| CN111833469A (en) * | 2019-04-18 | 2020-10-27 | 杭州海康威视数字技术股份有限公司 | Vehicle charging method and system applied to charging station |
| CN112002027A (en) * | 2020-09-04 | 2020-11-27 | 侯晓峰 | Highway charging method, device, terminal and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100261065B1 (en) * | 1997-12-30 | 2000-07-01 | 윤종용 | Vehicle classification method and apparatus using image analysis |
| JP2010134821A (en) * | 2008-12-08 | 2010-06-17 | Omron Corp | Vehicle-type discriminating apparatus |
| CN101783076A (en) * | 2010-02-04 | 2010-07-21 | 西安理工大学 | Method for quick vehicle type recognition under video monitoring mode |
| CN102637362A (en) * | 2012-04-01 | 2012-08-15 | 长安大学 | Tunnel vehicle type identification method based on video |
| CN103020582A (en) * | 2012-09-20 | 2013-04-03 | 苏州两江科技有限公司 | Method for computer to identify vehicle type by video image |
-
2013
- 2013-07-30 CN CN201310326455.8A patent/CN103413439B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100261065B1 (en) * | 1997-12-30 | 2000-07-01 | 윤종용 | Vehicle classification method and apparatus using image analysis |
| JP2010134821A (en) * | 2008-12-08 | 2010-06-17 | Omron Corp | Vehicle-type discriminating apparatus |
| CN101783076A (en) * | 2010-02-04 | 2010-07-21 | 西安理工大学 | Method for quick vehicle type recognition under video monitoring mode |
| CN102637362A (en) * | 2012-04-01 | 2012-08-15 | 长安大学 | Tunnel vehicle type identification method based on video |
| CN103020582A (en) * | 2012-09-20 | 2013-04-03 | 苏州两江科技有限公司 | Method for computer to identify vehicle type by video image |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104574986A (en) * | 2014-12-22 | 2015-04-29 | 贵州中南交通科技有限公司 | Vehicle model identifier with radiating device |
| CN105702050A (en) * | 2016-04-22 | 2016-06-22 | 安徽皖通科技股份有限公司 | Highway over-limit and overload management control method |
| CN108417042A (en) * | 2017-09-18 | 2018-08-17 | 西安金路交通工程科技发展有限责任公司 | Car based on vehicle image and lorry sorting technique |
| CN108171982A (en) * | 2018-02-06 | 2018-06-15 | 邓莉 | Crossing vehicular traffic vehicle monitor supervision platform |
| CN111833469A (en) * | 2019-04-18 | 2020-10-27 | 杭州海康威视数字技术股份有限公司 | Vehicle charging method and system applied to charging station |
| CN111145365A (en) * | 2019-12-17 | 2020-05-12 | 北京明略软件***有限公司 | Method, device, computer storage medium and terminal for realizing classification processing |
| CN112002027A (en) * | 2020-09-04 | 2020-11-27 | 侯晓峰 | Highway charging method, device, terminal and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103413439B (en) | 2016-01-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103413439B (en) | A kind of passenger vehicle based on video and lorry sorting technique | |
| TWI409718B (en) | Method of locating license plate of moving vehicle | |
| CN102722704B (en) | Method and system for recognizing vehicle license plate by integrating video dynamic tracking | |
| CN103279756B (en) | Vehicle detection based on integrated classifier analyzes system and determination method thereof | |
| CN102110376B (en) | Roadside parking space detection device based on computer vision | |
| CN102231236B (en) | Method and device for counting vehicles | |
| CN101510356B (en) | Video detection system and data processing device thereof, video detection method | |
| CN100573618C (en) | A kind of traffic intersection four-phase vehicle flow detection method | |
| CN102915433B (en) | Character combination-based license plate positioning and identifying method | |
| CN102867417B (en) | Taxi anti-forgery system and taxi anti-forgery method | |
| CN104732235A (en) | Vehicle detection method for eliminating night road reflective interference | |
| CN105203552A (en) | 360-degree tread image detecting system and method | |
| CN103049787A (en) | People counting method and system based on head and shoulder features | |
| CN103077387B (en) | Carriage of freight train automatic testing method in video | |
| CN105426894B (en) | Railway plug pin image detecting method and device | |
| CN104778444A (en) | Method for analyzing apparent characteristic of vehicle image in road scene | |
| CN103942560A (en) | High-resolution video vehicle detection method in intelligent traffic monitoring system | |
| EP2813973B1 (en) | Method and system for processing video image | |
| CN105336169A (en) | Method and system for judging traffic jams based on videos | |
| CN208476818U (en) | The dustiness detection system and vehicle washing system of rail transit rolling stock | |
| CN103337175A (en) | Vehicle type recognition system based on real-time video steam | |
| CN102789686B (en) | Road traffic flow detecting method based on road surface brightness composite mode recognition | |
| Hsia et al. | An Intelligent IoT-based Vision System for Nighttime Vehicle Detection and Energy Saving. | |
| CN112085018A (en) | License plate recognition system based on neural network | |
| CN103778790B (en) | A kind of vehicle flowrate square wave statistic law based on video sequence |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20160120 Termination date: 20170730 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |