CN108254738A - Obstacle-avoidance warning method, device and storage medium - Google Patents
Obstacle-avoidance warning method, device and storage medium Download PDFInfo
- Publication number
- CN108254738A CN108254738A CN201810098688.XA CN201810098688A CN108254738A CN 108254738 A CN108254738 A CN 108254738A CN 201810098688 A CN201810098688 A CN 201810098688A CN 108254738 A CN108254738 A CN 108254738A
- Authority
- CN
- China
- Prior art keywords
- thermal
- image
- data
- determined
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S11/00—Systems for determining distance or velocity not using reflection or reradiation
- G01S11/12—Systems for determining distance or velocity not using reflection or reradiation using electromagnetic waves other than radio waves
Landscapes
- Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Image Processing (AREA)
Abstract
This disclosure relates to a kind of obstacle-avoidance warning method, device and storage medium, it can be by obtaining the first thermal-induced imagery and the second thermal-induced imagery of barrier, first thermal-induced imagery and the second infrared chart seem to be obtained respectively by two thermal infrared imagers of different shooting angles, determine the parallax of first thermal-induced imagery and the second thermal-induced imagery, the distance of the barrier is determined according to the parallax, and when the distance is less than the first predetermined threshold value, obstacle-avoidance warning prompting is sent out.
Description
Technical field
This disclosure relates to intelligent distance-measuring avoidance field, and in particular, to a kind of obstacle-avoidance warning method, device and storage are situated between
Matter.
Background technology
With the continuous development of the technologies such as artificial intelligence, radar, computer vision, intelligent distance-measuring avoidance technology is also by more
Carry out the concern of more people.At present, intelligent distance-measuring avoidance technology is mainly used in robot, unmanned plane, diving apparatus, automatic Pilot
In the equipment such as vehicle or aircraft.
Existing intelligent distance-measuring technology divides active depth measuring (such as radar wave ranging, laser signal ranging) and passive type to survey
Two kinds away from (such as visible ligh-ranging), but active depth measuring technology is affected by environment larger, and easily detected counterattack, therefore concealment
Difference;Passive ranging relies primarily on visible ligh-ranging, and in dark surrounds, the mode of this visible ligh-ranging is relatively difficult to achieve, because
This does not adapt to round-the-clock ranging.
Invention content
To overcome problems of the prior art, the disclosure provides a kind of obstacle-avoidance warning method, device and storage medium.
According to the embodiment of the present disclosure in a first aspect, provide a kind of obstacle-avoidance warning method, the method includes:Obtain obstacle
The first thermal-induced imagery and the second thermal-induced imagery of object, first thermal-induced imagery and the second infrared chart seem by difference
Two thermal infrared imagers of shooting angle obtain respectively;Determine regarding for first thermal-induced imagery and the second thermal-induced imagery
Difference;The distance of the barrier is determined according to the parallax;When the distance is less than the first predetermined threshold value, obstacle-avoidance warning is sent out
Prompting.
Optionally, the parallax for determining first thermal-induced imagery and the second thermal-induced imagery includes:It determines described
The matching value of first thermal-induced imagery and the second thermal-induced imagery;The parallax is determined according to the matching value.
Optionally, the matching value for determining first thermal-induced imagery and the second thermal-induced imagery includes:Described
A wherein image is determined in first thermal-induced imagery and the second thermal-induced imagery as matching image, another image is Prototype drawing
Picture;Calculate the first mean value of the matching image and the second mean value of the template image;Described in being calculated by the following formula
Matching value:
Wherein, add_image represents the images match value after the matching image and template image superposition, image
Represent the pixel value that the matching image is each put, image_averageiRepresent matching images match window the i-th row picture
First mean value of element value, temp represent the pixel value that the template image is each put, temp_averageiRepresent the Prototype drawing
As the second mean value of the i-th row pixel value of match window.
Optionally, it is described to determine that the parallax includes according to the matching value:Minimum value in the matching value is determined
For data to be determined;It determines whether the data to be determined are valid data, and is determining that the data to be determined are significant figure
According to when, it is the index position of the parallax to determine the corresponding index position of the data to be determined;According to the rope of the parallax
Draw parallax described in location determination.
Optionally, described before determining whether the data to be determined are valid data, the method further includes:It determines
The gradient of the template image and;It is described to determine whether the data to be determined are that valid data include:It determines described to be determined
Whether whether data are less than the second predetermined threshold value, and the gradient corresponding with the data to be determined and preset more than third
Threshold value;It is less than the second predetermined threshold value in the data to be determined, and the gradient corresponding with the data to be determined and greatly
When third predetermined threshold value, determine that the data to be determined are valid data.
According to the second aspect of the embodiment of the present disclosure, a kind of obstacle-avoidance warning device is provided, described device includes:Obtain mould
Block, for obtaining the first thermal-induced imagery of barrier and the second thermal-induced imagery, first thermal-induced imagery and second red
Outer thermal image is obtained respectively by two thermal infrared imagers of different shooting angles;First determining module, for determining described
The parallax of one thermal-induced imagery and the second thermal-induced imagery;Second determining module, for determining the obstacle according to the parallax
The distance of object;Alarm module, for when the distance is less than the first predetermined threshold value, sending out obstacle-avoidance warning prompting.
Optionally, first determining module includes:First determination sub-module, for determining first thermal-induced imagery
And second thermal-induced imagery matching value;Second determination sub-module, for determining the parallax according to the matching value.
Optionally, first determination sub-module is used in first thermal-induced imagery and the second thermal-induced imagery really
A fixed wherein image is matching image, and another image is template image;Calculate it is described matching image the first mean value and
Second mean value of the template image;The matching value is calculated by the following formula:
Wherein, add_image represents the images match value after the matching image and template image superposition, image
Represent the pixel value that the matching image is each put, image_averageiRepresent matching images match window the i-th row picture
First mean value of element value, temp represent the pixel value that the template image is each put, temp_averageiRepresent the Prototype drawing
As the second mean value of the i-th row pixel value of match window.
Optionally, second determination sub-module is used to the minimum value in the matching value being determined as data to be determined;
It determines whether the data to be determined are valid data, and when it is valid data to determine the data to be determined, determines described
The corresponding index position of data to be determined is the index position of the parallax;According to determining the index position of the parallax
Parallax.
Optionally, described device further includes:Third determining module, for determine the gradient of the template image and;It is described
Second determination sub-module for determine the data to be determined whether be less than the second predetermined threshold value, and with the data to be determined
The corresponding gradient and whether be more than third predetermined threshold value;In the data to be determined less than the second predetermined threshold value, and with
The corresponding gradient of the data to be determined and during more than third predetermined threshold value, determines that the data to be determined are significant figures
According to.
According to the third aspect of the embodiment of the present disclosure, a kind of obstacle-avoidance warning device is provided, described device includes:Processor;
For storing the memory of processor-executable instruction;Wherein, the processor is used to obtain the first infrared chart of barrier
As and the second thermal-induced imagery, first thermal-induced imagery and the second infrared chart seem red by two of different shooting angles
Outer thermal imaging system obtains respectively;Determine the parallax of first thermal-induced imagery and the second thermal-induced imagery;It is true according to the parallax
The distance of the fixed barrier;When the distance is less than the first predetermined threshold value, obstacle-avoidance warning prompting is sent out.
According to the fourth aspect of the embodiment of the present disclosure, a kind of computer readable storage medium is provided, is stored thereon with calculating
The step of machine program instruction, described program instruction realizes the disclosure first aspect the method when being executed by processor.
It is described by obtaining the first thermal-induced imagery and the second thermal-induced imagery of barrier using above-mentioned technical proposal
First thermal-induced imagery and the second infrared chart seem to be obtained respectively by two thermal infrared imagers of different shooting angles;Determine institute
State the parallax of the first thermal-induced imagery and the second thermal-induced imagery;The distance of the barrier is determined according to the parallax;Institute
When stating distance less than the first predetermined threshold value, obstacle-avoidance warning prompting is sent out, the scheme that the disclosure provides is not needed to using a large amount of precisions
Mechanical scanning component saves cost, can also realize and be included in that night, smog, acutely concussion etc. is round-the-clock under mal-conditions
Barrier real time distance, and concealment is stronger, detected counterattack is not easy, so as to better meet the demand of user.
Other feature and advantage of the disclosure will be described in detail in subsequent specific embodiment part.
Description of the drawings
Attached drawing is for providing further understanding of the disclosure, and a part for constitution instruction, with following tool
Body embodiment is used to explain the disclosure, but do not form the limitation to the disclosure together.In the accompanying drawings:
Fig. 1 is the flow chart according to a kind of obstacle-avoidance warning method shown in an exemplary embodiment;
Fig. 2 is the block diagram according to the first obstacle-avoidance warning device shown in an exemplary embodiment;
Fig. 3 is the block diagram according to second of obstacle-avoidance warning device shown in an exemplary embodiment;
Fig. 4 is the block diagram according to the third obstacle-avoidance warning device shown in an exemplary embodiment.
Specific embodiment
The specific embodiment of the disclosure is described in detail below in conjunction with attached drawing.It should be understood that this place is retouched
The specific embodiment stated is only used for describing and explaining the disclosure, is not limited to the disclosure.
The disclosure can be applied to the scene of intelligent distance-measuring avoidance, such as robot, unmanned plane, diving apparatus, vehicle pair
The timely alarm of barrier, in this scenario, existing intelligent distance-measuring technology are broadly divided into active depth measuring and passive ranging
Two kinds, wherein, active depth measuring is mainly using radar wave or laser signal ranging, by establishing sender unit and connecing
Correspondence between receiving apparatus using the transmitting feedback information of signal, measures a kind of method of target range;Passive ranging
Dedicated sender unit is not needed to, but by the analyzing and processing to visual signal come mensuration distance, still, active survey
Away from method it is affected by environment larger, for example, during illumination abundance, laser is interfered by ambient noise and generates range error, so as to
Lead to the avoid-obstacle behavior of mistake, in addition, the method for active depth measuring needs actively to emit target searchlighting signal, it is easily detected anti-
Hit, therefore concealment is poor, also, active depth measuring needs a large amount of precision machinery sweep units, this can not only cause cost compared with
Height, real-time are also slightly worse;Existing passive ranging relies primarily on visible ligh-ranging, in dark surrounds, this visible flash ranging
Away from mode effect it is unsatisfactory, therefore do not adapt to round-the-clock ranging.
To solve problems of the prior art, present disclose provides a kind of obstacle-avoidance warning method, device and storages to be situated between
Matter using infrared thermal imaging technique, binocular distance measurement technology and straight-line detection technology, is detecting front obstacle distance
During less than predetermined threshold value, obstacle-avoidance warning prompting is sent out, wherein, infrared thermal imaging technique is a kind of passive infrared night vision technology,
Principle is to be higher than the object of absolute zero (- 273 DEG C) based on all temperature in nature, all gives off infrared ray all the time,
This infrared radiation simultaneously is all loaded with the characteristic information of object, this is just the temperature that various measured targets are differentiated using infrared technique
Degree height and heat distribution field provide objective basis;Binocular distance measurement technology is the passive ranging established according to principle of parallax
System, i.e., when observing same barrier using images of left and right eyes, due to barrier, image space is different and exist and regard in right and left eyes
The distance of barrier can be calculated using this parallax for difference, so as to when determining that the distance is less than the first predetermined threshold value, send out
Obstacle-avoidance warning is prompted.
The disclosure is described in detail below by specific embodiment.
Fig. 1 is according to a kind of flow chart of obstacle-avoidance warning method shown in an exemplary embodiment, as shown in Figure 1, at this
In embodiment, illustrate, include the following steps by taking the scene of unmanned plane avoidance as an example:
In a step 101, the first thermal-induced imagery and the second thermal-induced imagery of barrier are obtained.
Wherein, which can include high building, control tower, high-voltage line etc., first thermal-induced imagery and the second infrared heat
Image is obtained respectively by two thermal infrared imagers of different shooting angles, wherein, which is located at same water
Square upwards, in a kind of possible realization method, can two thermal infrared imagers be individually positioned in by a fixation by bolt
The left and right sides of frame, and each placement location of the thermal infrared imager on the fixed frame can change, and which exists
After the placement location of fixed frame determines, that is, the measurement parameter (such as measurement range, measurement accuracy) of the thermal infrared imager is determined,
In this way, the first thermal-induced imagery and the second thermal-induced imagery of the barrier can be obtained by the two thermal infrared imagers.
In a step 102, the parallax of first thermal-induced imagery and the second thermal-induced imagery is determined.
The disclosure is to determine the barrier and unmanned plane current location based on the principle of parallax in binocular distance measurement technology
Distance, the principle of parallax refer to using images of left and right eyes observe same barrier when, due to barrier in right and left eyes image space
It is different and there are parallax, the distance of barrier can be calculated using this parallax, two thermal infrared imagers in the disclosure can
Using the simulator as images of left and right eyes, the thermal-induced imagery of same barrier is obtained from different perspectives respectively, in the disclosure
In, which refers to after being superimposed first thermal-induced imagery and second thermal-induced imagery, the same spy on same barrier
The difference of the relative position of subpoint of the sign o'clock on two thermal-induced imageries, for example, setting same characteristic point on barrier as A
Point, subpoint of the A points on first thermal-induced imagery are A1Point, subpoint of the A points on second thermal-induced imagery are A2
Point, at this point, the parallax can be A1Point and A2The distance of point, above-mentioned to be merely illustrative, the disclosure is not construed as limiting this.
It in this step, can be by carrying out images match to first thermal-induced imagery and the second thermal-induced imagery, really
The matching value of fixed first thermal-induced imagery and the second thermal-induced imagery, then determines the parallax according to the matching value.
It, can be in first infrared chart in the matching value for determining first thermal-induced imagery and the second thermal-induced imagery
A wherein image is determined in picture and the second thermal-induced imagery as matching image, another image is template image, is then calculated
First mean value of the matching image and the second mean value of the template image, and pass through the following formula and calculate the matching value:
Wherein, add_image represents the images match value after the matching image and template image superposition, image
Represent the pixel value that the matching image is each put, image_averageiRepresent matching images match window the i-th row picture
First mean value of element value, temp represent the pixel value that the template image is each put, temp_averageiRepresent the Prototype drawing
As the second mean value of the i-th row pixel value of match window.
When carrying out images match, a kind of possible realization method can be matched using the mode of match window,
For example, the size of matching image and template image is all X*Y, and after matching effect and resource occupation problem is considered, choosing
The match window for selecting 9*9 is matched, and to realize the matching of X*Y images, the matching of 9*9 images is first carried out, to realize 9*9 images
Matching, the matching (image data be by row transmit) of 1*9 images is first carried out, when carrying out images match, with first 1*9 figure
For the matching of picture, the point of the expression location of pixels of first 1*9 matrix of template image is 1,2,3 ... 9, match image first
The point of the expression location of pixels of a 1*9 matrixes is also 1,2,3 ... 9, at this point, keeping template image constant, moved after matching image
One, i.e., 2,3,4 ... 10,64,65,66 are moved to when matching image ... when 72, shifting one after the point of template image, i.e., 2,
3,4 ... 10, matching image again with point 2,3,4 ... 10 for starting, and so on can obtain (X- in a line 1*9 matrixes
72) * 64 matching values, above-mentioned example are merely illustrative, and the disclosure is not construed as limiting this.
Before the matching value is calculated, also need to calculate the first mean value of matching images match i-th row pixel value of window
image_averageiAnd the second mean value temp_average of i-th row pixel value of template image match windowi, in one kind
In possible realization method, can this be calculated by formula add_image=add_image+image+image_reg first
Superposition of data with image and the template image and, wherein, the add_image on the equal sign left side is the superposition that this will be calculated
Data and, the add_image on the right of equal sign be the last superposition of data calculated adjacent with this calculating and, image can be with
It is the pixel value of image, image_reg is the image pixel value with register cache, then should according to the superposition of data and calculating
First mean value and the second mean value, it is contemplated that image data is transmitted by row, therefore, can calculate every a line image respectively
The superposition of data of data and, and with the superposition of data and divided by the every number of a line image pixel point, you can obtain every a line figure
As first mean value of data and the second mean value, for example, for the Window match mode of 9*9, each 1*9 matrixes can be calculated
Superposition of data and, the superposition of data that is exported after 8 operations are carried out and be add_image=add_image+image+
Image_reg-image_0, image_0 are first number in a upper 1*9 matrix, for the image of X*Y sizes, successively into
Row operation can obtain the first mean value image_average of X-8 1*9 matrixiAnd the second mean value temp_averagei, and
The the first mean value image_average that will be obtained using memoryiAnd the second mean value temp_averageiIt is stored, example
Such as, for ease of calling the data of storage in order, FIFO (First Input First Output) first in, first out can be utilized
Queue memory stores first mean value and the second mean value, and the data being introduced into the FIFO memory are first called and retire from office,
And then second data is just called, of first thermal-induced imagery and the second thermal-induced imagery is finally calculated using formula
With value.
When determining the parallax according to the matching value, the minimum value in the matching value can be determined as data to be determined,
Then determine whether the data to be determined are valid data, and when it is valid data to determine the data to be determined, determine that this is treated
It is the index position of the parallax to determine the corresponding index position of data;The parallax is determined according to the index position of the parallax,
In, which refers to the position of subpoint of the same characteristic point on two thermal-induced imageries on barrier, for example, being
Convenient for calculating, it is 0 that can preset the position of corresponding mapping point on template image, then in matching value minimum value position, that is, parallax
Index position, according to the index position of the parallax i.e. can determine the parallax.
Before determining whether the data to be determined are valid data, also need to determine the template image gradient and, pass through
It determines whether the data to be determined meet preset condition, and when the data to be determined meet the preset condition, determines that this is treated really
Fixed number evidence is valid data, wherein, which can include:The data to be determined be less than the second predetermined threshold value, and with
The corresponding gradient of the data to be determined and more than third predetermined threshold value.
Calculation template image gradient and when, using with calculating the first mean value and an equal amount of matching window of the second mean value
Mouthful, for example, the match window of 9*9 can be selected, then pass through formula add_gradient=add_gradient+image-
Image_reg calculate the template image data and, wherein, the add_gradient on the equal sign left side is the ladder that this will be calculated
Degree and, the add_gradient on the right of equal sign be the last gradient calculated adjacent with this calculating and, image can be figure
The pixel value of picture, image_reg is the image data with register cache, in a kind of possible realization method, it also is contemplated that
To image data be by row be transmitted, therefore, can calculate respectively every a line image data gradient and, for example, for 9*9
Window match mode, can calculate each 1*9 matrixes gradient and, the summed data exported after 8 operations are carried out is
Add_gradient=add_gradient+image+image_reg-image_0, wherein, image_0 is and this calculating phase
The second pixel value of an adjacent upper 1*9 matrix and the difference of first pixel value, can be obtained by carrying out operation successively by X-8
The gradient and data of 1*9 matrixes, and being stored obtained data using FIFO memory, be more than when the line number of storage or
During equal to 8 row, read the data of FIFO memory, then calculate 9*9 images gradient and, then the data of storage are passed
It pushing away, i.e., first data is stored in second FIFO, in second data deposit third FIFO, and so on, realize 9*9 figures
As pressing row mobile computing.
It should be noted that whether the gradient and the characteristic of image that can be used for judging to get belong to low line
Reason, for example, the infrared picture data of a face wall or sky belongs to low texture, the image data of low texture carrys out mensuration distance
It says, without reference value, when determining the gradient and less than or equal to the third predetermined threshold value, it is believed that get
Image data is invalid data, in addition, by the formula of calculating matching value it is found that the matching value is matching image data and Prototype drawing
As the absolute value of data difference, which shows that the matching image data corresponds to reflecting for the characteristic point of barrier closer to 0
The mapping point 2 of the characteristic point of the barrier corresponding with template image data of exit point 1 is closer, so can consider the mapping point 1 with
What mapping point 2 represented is the same characteristic point on barrier, therefore, second predetermined threshold value is pre-set, in the number to be determined
According to less than second predetermined threshold value, and gradient corresponding with the data to be determined and the two are pre- more than the third predetermined threshold value
If during condition, it may be determined that the data to be determined are valid data, and then can determine the parallax.
Furthermore, it is contemplated that the real-time of operation, the disclosure uses on-site programmable gate array FPGA (Field-
Programmable Gate Array) and ARM (Advanced RISC Machine) processor framework for being combined realize
Parallax is calculated in real time as a result, and using arm processor calculating distance, the image processing architecture based on fpga chip can be realized more
Road parallel computation has high real-time, high reliability, low-power consumption compared to other implementations based on x86, GPU, ARM framework
Etc. advantage, the characteristics of due to FPGA concurrent operations, the calculating of subsequent disparity correspondence value and mean data and gradient and
Calculating can synchronize progress.
So that matching value is more accurate, before images match is carried out, also need to first thermal-induced imagery and this
Two thermal-induced imageries carry out image preprocessing, in a kind of possible realization method, to first thermal-induced imagery and can be somebody's turn to do
Second thermal-induced imagery carries out the pretreatment operations such as sequential alignment, target area interception, image rectification, can pass through HDMI cable
(can also be required according to technology be extended the communication modes such as RS422, RS485h and HDMI high definition inputted come response external) is by two
The Digital Image Transmission of a infrared thermal imaging carries out image pretreatment operation to image processing module.
In the disclosure, since first thermal-induced imagery and second infrared chart seem by the two of different shooting angles
A thermal infrared imager obtains respectively, and two thermal infrared imagers are two self-contained units, it is therefore desirable to by the first infrared heat
Image and the sequential of second thermal-induced imagery input are aligned, and the calculating with images match value is corrected convenient for subsequent image,
Illustratively, using the input timing of the first thermal-induced imagery as standard, by the image data of the second thermal-induced imagery with parity field
Mode stores in the memory unit, and the data of storage unit are then read according to the input timing of the first thermal-induced imagery, realizes
Synchrodata exports, and above-mentioned example is merely illustrative, and the disclosure is not construed as limiting this.
After first thermal-induced imagery and second thermal-induced imagery are carried out sequential alignment, to reduce the complexity calculated
Degree also needs to carry out target area interception to two thermal-induced imagery data, target area can be specifically realized by way of compression
The interception in domain, compress mode can be the modes such as image sampling, image local interception, image array averaging, in a kind of possibility
Realization method in, can the first thermal-induced imagery after alignment and the second thermal-induced imagery be inputted into two target areas respectively
Interception module carries out target area interception, for example, for the thermal-induced imagery of a 32*32 size, it can be carried out in " field "
Stroke point is divided into the thermal-induced imagery of 4 16*16 sizes, and the thermal-induced imagery of every 16*16 size is the mesh intercepted
Logo image, when subsequently carrying out matching primitives, the thermal-induced imagery of every 16*16 size can be carried out respectively image rectification and
Matching primitives reduce the complexity of calculating, improve the real-time of system response.
It is after target area interception is carried out to image, first thermal-induced imagery and second thermal-induced imagery is corresponding
Target interception image is transferred to two image correction modules and carries out image rectification respectively, and specific implementation can cut target
Image is taken to be respectively stored in two storage units, and pre-set correction parameter is stored in another storage unit,
The correction parameter can include reading target interception image access unit address and correction coefficient, for example, the correction parameter
Can be the data of 64, for the image of X*Y sizes, the maximum value of the address in the correction parameter is X*Y, and correction coefficient can
To be x1, x2, x3, tetra- parameters of x4, when being corrected to image, address information first in correction parameter is from target
4 data are read in interception image, and this 4 data correspond to the image pixel value in 2*2 regions in target interception image respectively,
For ease of narration, this 4 data are represented respectively with a, b, c, d, and the address of this 4 data can use Address respectively,
Address+1, Address+Y, Address+Y+1 represent, wherein, Address represents the address of data a, Address+1 tables
Show with data a positioned at the address with a line and adjacent with data a data b, Address+Y represents to be located at data a same
Row and the data c adjacent with data a address, Address+Y+1 represent with data c positioned at a line and with data c phases
The address of adjacent data d, for example, for the image of a 3*3, image pixel value isIt can pre-set
The address matrix of the 3*3 images isFor Address=0, Address+1=1, Address+Y=3,
For this four addresses of Address+Y+1=4, corresponding 4 data areIt is merely illustrative herein, the disclosure
This is not construed as limiting, then obtain 4 data are multiplied respectively with 4 correction coefficient in correction parameter, and sums and obtains
Image data after correction, i.e. I=I (Address) * x1+I (Address+1) * x2+I (Address+Y) * x3+I (Address
+ Y+1) * x4, wherein, I (Address) represents the corresponding image pixel values of address Address, above-mentioned to be merely illustrative, this public affairs
It opens and this is not construed as limiting.It should be noted that correction parameter is constant after calibration early period, so when each frame image data
It is corrected into fashionable, realizes the real-time calculating of correction.
After image completes correction, the image for correcting completion can be input to images match module and carry out images match,
So as to which the parallax of first thermal-induced imagery and second thermal-induced imagery be calculated.
In step 103, the distance of the barrier is determined according to the parallax.
In this step, the distance of the barrier can be calculated by the following formula Z=fT/d, wherein Z represents barrier
Hinder the distance of object, f represents the focal length of camera lens used in thermal infrared imager, and T represents baseline length, i.e. between two thermal infrared imagers
Distance, d represent parallax.
At step 104, when the distance is less than the first predetermined threshold value, obstacle-avoidance warning prompting is sent out.
It in this step, can be to next system (such as aircraft control system when determining that the distance is less than the first predetermined threshold value
System) it sends avoidance instruction message or is connected on acoustical generator, avoiding barrier is issued the user with by way of audio alarm
Alarm prompt, then the image information of barrier is directly transferred to image display, and the image information subscript
It is marked with the distance of the barrier so that user can more intuitively see the image information of front obstacle on display interface.
In a kind of possible realization method, which is the standard using fpga chip generation pal mode
Sequential, and depth image is generated under thermal imaging system coordinate system according to 25Hz frame frequencies, which can also be referred to as distance map
Picture refers to the distance (depth) using each point on from thermal imaging system to barrier as the image of pixel value, it directly reflects obstacle
The geometry of object visible surface.Different display patterns can also be configured in the image display, for example, five kinds can be configured
Display pattern shows the first thermal-induced imagery, the second thermal-induced imagery, the first thermal-induced imagery and the second thermal-induced imagery respectively
Superimposed image, range image and preceding four kinds of image sets are included into the multiwindow image in an image, it is above-mentioned only to lift
Example explanation, the disclosure are not construed as limiting this.
In addition, when the barrier in the thermal-induced imagery got belongs to line style barrier (such as high-tension bus-bar), need
The coordinate value of the line style barrier is obtained by the method for straight-line detection, the line style barrier is then determined according to the coordinate value
Distance.
Illustratively, the coordinate value can be calculated by sobel operators, specific calculating process belongs to the prior art, herein
It repeats no more.
In addition, the obstacle-avoidance warning method provided using the disclosure, under the mal-conditions such as night, smog, violent concussion
The distance of barrier can be detected, and when the distance is less than predetermined threshold value, send out obstacle-avoidance warning prompting, the avoidance that the disclosure provides
The visibility of alarm device is 4km-10km, barrier that can be in the range of real time distance 300m-950m.
Using the above method, do not need to, using a large amount of precision machinery sweep units, save cost, can also realize and be included in
It is round-the-clock to barrier progress real time distance under the mal-conditions such as night, smog, violent concussion, and concealment is stronger,
Detected counterattack is not easy, so as to better meet the demand of user.
Fig. 2 is according to a kind of block diagram of obstacle-avoidance warning device shown in an exemplary embodiment, as shown in Fig. 2, the device
Including:
Acquisition module 201, for obtaining the first thermal-induced imagery of barrier and the second thermal-induced imagery, this is first infrared
Thermal image and the second infrared chart seem to be obtained respectively by two thermal infrared imagers of different shooting angles;
First determining module 202, for determining the parallax of first thermal-induced imagery and the second thermal-induced imagery;
Second determining module 203, for determining the distance of the barrier according to the parallax;
Alarm module 204, for when the distance is less than the first predetermined threshold value, sending out obstacle-avoidance warning prompting.
Optionally, Fig. 3 is a kind of block diagram of obstacle-avoidance warning device shown in illustrated embodiment according to fig. 2, as shown in figure 3,
First determining module 202 includes:
First determination sub-module 2021, for determining the matching value of first thermal-induced imagery and the second thermal-induced imagery;
Second determination sub-module 2022, for determining the parallax according to the matching value.
Optionally, which is used in first thermal-induced imagery and the second thermal-induced imagery really
A fixed wherein image is matching image, and another image is template image;It calculates the first mean value of the matching image and is somebody's turn to do
Second mean value of template image;The matching value is calculated by the following formula:
Wherein, add_image represents the images match value after the matching image and template image superposition, image
Represent the pixel value that the matching image is each put, image_averageiRepresent matching images match window the i-th row picture
First mean value of element value, temp represent the pixel value that the template image is each put, temp_averageiRepresent the Prototype drawing
As the second mean value of the i-th row pixel value of match window.
Optionally, which is used to the minimum value in the matching value being determined as data to be determined;
It determines whether the data to be determined are valid data, and when it is valid data to determine the data to be determined, determines that this is to be determined
The corresponding index position of data is the index position of the parallax;The parallax is determined according to the index position of the parallax.
Optionally, Fig. 4 is a kind of block diagram of obstacle-avoidance warning device shown in illustrated embodiment according to fig. 3, as shown in figure 4,
The device further includes:
Third determining module 205, for determine the gradient of the template image and;
Second determination sub-module 2022 determines whether the data to be determined are less than the second predetermined threshold value, and treat really with this
Fixed number is according to the corresponding gradient and whether is more than third predetermined threshold value, is less than the second predetermined threshold value in the data to be determined, and
Gradient corresponding with the data to be determined and during more than third predetermined threshold value, determines that the data to be determined are valid data.
Using above device, do not need to, using a large amount of precision machinery sweep units, save cost, can also realize and be included in
Night, smog, the acutely round-the-clock barrier real time distance under mal-conditions such as concussion, and concealment is stronger, be not easy by
Detection counterattack, so as to better meet the demand of user.
The disclosure also provides a kind of obstacle-avoidance warning device, which includes:Processor;Finger is can perform for storing processor
The memory of order;Wherein, which is used to obtain the first thermal-induced imagery and the second thermal-induced imagery of barrier, this first
Thermal-induced imagery and the second infrared chart seem to be obtained respectively by two thermal infrared imagers of different shooting angles;Determine this first
The parallax of thermal-induced imagery and the second thermal-induced imagery;The distance of the barrier is determined according to the parallax;It is less than the in the distance
During one predetermined threshold value, obstacle-avoidance warning prompting is sent out.
The disclosure also provides a kind of computer readable storage medium, is stored thereon with computer program instructions, which refers to
The step of above-mentioned obstacle-avoidance warning method is realized when order is executed by processor.
The preferred embodiment of the disclosure is described in detail above in association with attached drawing, still, the disclosure is not limited to above-mentioned reality
The detail in mode is applied, in the range of the technology design of the disclosure, a variety of letters can be carried out to the technical solution of the disclosure
Monotropic type, these simple variants belong to the protection domain of the disclosure.
It is further to note that specific technical features described in the above specific embodiments, in not lance
In the case of shield, can be combined by any suitable means, in order to avoid unnecessary repetition, the disclosure to it is various can
The combination of energy no longer separately illustrates.
In addition, arbitrary combination can also be carried out between a variety of different embodiments of the disclosure, as long as it is without prejudice to originally
Disclosed thought should equally be considered as disclosure disclosure of that.
Claims (12)
1. a kind of obstacle-avoidance warning method, which is characterized in that the method includes:
Obtain the first thermal-induced imagery and the second thermal-induced imagery of barrier, first thermal-induced imagery and the second infrared heat
Image is obtained respectively by two thermal infrared imagers of different shooting angles;
Determine the parallax of first thermal-induced imagery and the second thermal-induced imagery;
The distance of the barrier is determined according to the parallax;
When the distance is less than the first predetermined threshold value, obstacle-avoidance warning prompting is sent out.
2. according to the method described in claim 1, it is characterized in that, first thermal-induced imagery and second infrared of determining
The parallax of thermal image includes:
Determine the matching value of first thermal-induced imagery and the second thermal-induced imagery;
The parallax is determined according to the matching value.
3. according to the method described in claim 2, it is characterized in that, first thermal-induced imagery and second infrared of determining
The matching value of thermal image includes:
A wherein image is determined in first thermal-induced imagery and the second thermal-induced imagery as matching image, another figure
As being template image;
Calculate the first mean value of the matching image and the second mean value of the template image;
The matching value is calculated by the following formula:
Wherein, add_image represents the images match value after the matching image and template image superposition, and image is represented
The pixel value that the matching image is each put, image_averageiRepresent matching images match i-th row pixel value of window
The first mean value, temp represents the pixel value that the template image is each put, temp_averageiRepresent the template image
The second mean value with the i-th row pixel value of window.
4. according to the method described in claim 2, it is characterized in that, described determine that the parallax includes according to the matching value:
Minimum value in the matching value is determined as data to be determined;
It determines whether the data to be determined are valid data, and when it is valid data to determine the data to be determined, determines
The corresponding index position of the data to be determined is the index position of the parallax;
The parallax is determined according to the index position of the parallax.
5. method according to claim 3 or 4, which is characterized in that described to determine whether the data to be determined are to have
Before imitating data, the method further includes:
Determine the template image gradient and;
It is described to determine whether the data to be determined are that valid data include:
Determine whether the data to be determined are less than the second predetermined threshold value, and the gradient corresponding with the data to be determined
Whether third predetermined threshold value is more than;
It is less than the second predetermined threshold value in the data to be determined, and the gradient corresponding with the data to be determined and is more than
During third predetermined threshold value, determine that the data to be determined are valid data.
6. a kind of obstacle-avoidance warning device, which is characterized in that described device includes:
Acquisition module, for obtaining the first thermal-induced imagery of barrier and the second thermal-induced imagery, first infrared chart
Picture and the second infrared chart seem to be obtained respectively by two thermal infrared imagers of different shooting angles;
First determining module, for determining the parallax of first thermal-induced imagery and the second thermal-induced imagery;
Second determining module, for determining the distance of the barrier according to the parallax;
Alarm module, for when the distance is less than the first predetermined threshold value, sending out obstacle-avoidance warning prompting.
7. device according to claim 6, which is characterized in that first determining module includes:
First determination sub-module, for determining the matching value of first thermal-induced imagery and the second thermal-induced imagery;
Second determination sub-module, for determining the parallax according to the matching value.
8. device according to claim 7, which is characterized in that first determination sub-module is used for infrared described first
A wherein image is determined in thermal image and the second thermal-induced imagery as matching image, another image is template image;It calculates
First mean value of the matching image and the second mean value of the template image;The matching value is calculated by the following formula:
Wherein, add_image represents the images match value after the matching image and template image superposition, and image is represented
The pixel value that the matching image is each put, image_averageiRepresent matching images match i-th row pixel value of window
The first mean value, temp represents the pixel value that the template image is each put, temp_averageiRepresent the template image
The second mean value with the i-th row pixel value of window.
9. device according to claim 7, which is characterized in that second determination sub-module is used for will be in the matching value
Minimum value be determined as data to be determined;Determine whether the data to be determined are valid data, and determine it is described to be determined
When data are valid data, it is the index position of the parallax to determine the corresponding index position of the data to be determined;According to
The index position of the parallax determines the parallax.
10. device according to claim 8 or claim 9, which is characterized in that described device further includes:
Third determining module, for determine the gradient of the template image and;
Second determination sub-module is treated for whether determining the data to be determined less than the second predetermined threshold value with described
It determines the corresponding gradient of data and whether is more than third predetermined threshold value;It is less than the second default threshold in the data to be determined
Value, and the gradient corresponding with the data to be determined and during more than third predetermined threshold value, determine the data to be determined
It is valid data.
11. a kind of obstacle-avoidance warning device, which is characterized in that described device includes:
Processor;
For storing the memory of processor-executable instruction;
Wherein, the processor is used to obtain the first thermal-induced imagery and the second thermal-induced imagery of barrier, and described first is red
Outer thermal image and the second infrared chart seem to be obtained respectively by two thermal infrared imagers of different shooting angles;Determine described first
The parallax of thermal-induced imagery and the second thermal-induced imagery;The distance of the barrier is determined according to the parallax;In the distance
During less than the first predetermined threshold value, obstacle-avoidance warning prompting is sent out.
12. a kind of computer readable storage medium, is stored thereon with computer program instructions, which is characterized in that described program refers to
The step of any one of claim 1 to 6 the method is realized when order is executed by processor.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810098688.XA CN108254738A (en) | 2018-01-31 | 2018-01-31 | Obstacle-avoidance warning method, device and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810098688.XA CN108254738A (en) | 2018-01-31 | 2018-01-31 | Obstacle-avoidance warning method, device and storage medium |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108254738A true CN108254738A (en) | 2018-07-06 |
Family
ID=62743558
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810098688.XA Pending CN108254738A (en) | 2018-01-31 | 2018-01-31 | Obstacle-avoidance warning method, device and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108254738A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109218674A (en) * | 2018-09-25 | 2019-01-15 | 杭州翼兔网络科技有限公司 | A kind of diving apparatus group control system |
CN110838138A (en) * | 2019-09-26 | 2020-02-25 | 北京迈格威科技有限公司 | Repetitive texture detection method, device, computer equipment and storage medium |
CN113834571A (en) * | 2020-06-24 | 2021-12-24 | 杭州海康威视数字技术股份有限公司 | Target temperature measurement method, device and temperature measurement system |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2012128314A1 (en) * | 2011-03-23 | 2012-09-27 | シャープ株式会社 | Parallax calculation device, distance calculation device and parallax calculation method |
CN103568990A (en) * | 2012-07-23 | 2014-02-12 | 北京新岸线移动多媒体技术有限公司 | Method and system for achieving vehicle safety warning |
WO2014030630A1 (en) * | 2012-08-20 | 2014-02-27 | 株式会社日本自動車部品総合研究所 | Device for generating disparity map and method therefor |
CN103679686A (en) * | 2012-09-11 | 2014-03-26 | 株式会社理光 | Match measure determination method, match measure determination device, parallax computation method and image matching method |
CN106444837A (en) * | 2016-10-17 | 2017-02-22 | 北京理工大学 | Obstacle avoiding method and obstacle avoiding system for unmanned aerial vehicle |
CN107403448A (en) * | 2017-07-26 | 2017-11-28 | 海信集团有限公司 | Cost function generation method and cost function generating means |
-
2018
- 2018-01-31 CN CN201810098688.XA patent/CN108254738A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2012128314A1 (en) * | 2011-03-23 | 2012-09-27 | シャープ株式会社 | Parallax calculation device, distance calculation device and parallax calculation method |
CN103568990A (en) * | 2012-07-23 | 2014-02-12 | 北京新岸线移动多媒体技术有限公司 | Method and system for achieving vehicle safety warning |
WO2014030630A1 (en) * | 2012-08-20 | 2014-02-27 | 株式会社日本自動車部品総合研究所 | Device for generating disparity map and method therefor |
CN103679686A (en) * | 2012-09-11 | 2014-03-26 | 株式会社理光 | Match measure determination method, match measure determination device, parallax computation method and image matching method |
CN106444837A (en) * | 2016-10-17 | 2017-02-22 | 北京理工大学 | Obstacle avoiding method and obstacle avoiding system for unmanned aerial vehicle |
CN107403448A (en) * | 2017-07-26 | 2017-11-28 | 海信集团有限公司 | Cost function generation method and cost function generating means |
Non-Patent Citations (3)
Title |
---|
YEAN YIN 等: ""A Stereo Pairs Disparity Matching Algorithm by Mean-Shift Segmentation"", 《THIRD INTERNATIONAL WORKSHOP ON ADVANCED COMPUTATIONAL INTELLIGENCE》 * |
郭宝龙 等: "《数字图像处理***工程讨论》", 31 July 2012 * |
顾征 等: ""采用邻域差值特征模板的立体匹配方法"", 《光电工程》 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109218674A (en) * | 2018-09-25 | 2019-01-15 | 杭州翼兔网络科技有限公司 | A kind of diving apparatus group control system |
CN110838138A (en) * | 2019-09-26 | 2020-02-25 | 北京迈格威科技有限公司 | Repetitive texture detection method, device, computer equipment and storage medium |
CN113834571A (en) * | 2020-06-24 | 2021-12-24 | 杭州海康威视数字技术股份有限公司 | Target temperature measurement method, device and temperature measurement system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112132972B (en) | Three-dimensional reconstruction method and system for fusing laser and image data | |
CN112634374B (en) | Stereoscopic calibration method, device and system for binocular camera and binocular camera | |
KR101697512B1 (en) | Image registration device and method thereof | |
US20110211045A1 (en) | Method and system for producing multi-view 3d visual contents | |
CN110009672A (en) | Promote ToF depth image processing method, 3D rendering imaging method and electronic equipment | |
CN108254738A (en) | Obstacle-avoidance warning method, device and storage medium | |
CN113358231B (en) | Infrared temperature measurement method, device and equipment | |
US9628777B2 (en) | Method of 3D reconstruction of a scene calling upon asynchronous sensors | |
CN104299231B (en) | Method and system for registering images of multiple sensors in real time | |
CN104794713A (en) | Greenhouse crop digital-imaging method based on ARM and binocular vision | |
CN110889873A (en) | Target positioning method and device, electronic equipment and storage medium | |
Zalud et al. | Fusion of thermal imaging and CCD camera-based data for stereovision visual telepresence | |
CN114526710A (en) | Sea surface measuring system, sea surface measuring method, and storage medium | |
CN110619660A (en) | Object positioning method and device, computer readable storage medium and robot | |
CN103528562A (en) | Method for detecting distance of human eyes and display terminal based on single camera | |
WO2018216341A1 (en) | Information processing device, information processing method, and program | |
CN107564051B (en) | Depth information acquisition method and system | |
CN111798507A (en) | Power transmission line safety distance measuring method, computer equipment and storage medium | |
Belhedi et al. | Noise modelling in time‐of‐flight sensors with application to depth noise removal and uncertainty estimation in three‐dimensional measurement | |
CN113724391A (en) | Three-dimensional model construction method and device, electronic equipment and computer readable medium | |
Hahne et al. | Depth imaging by combining time-of-flight and on-demand stereo | |
Shen et al. | Extrinsic calibration for wide-baseline RGB-D camera network | |
JP2011165081A (en) | Method and apparatus for generating image, and program | |
CN116704125B (en) | Mapping method, device, chip and module equipment based on three-dimensional point cloud | |
CN111742352A (en) | 3D object modeling method and related device and computer program product |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180706 |