CN106560861A - Intelligent supervision method based on computer vision - Google Patents
Intelligent supervision method based on computer vision Download PDFInfo
- Publication number
- CN106560861A CN106560861A CN201510635826.XA CN201510635826A CN106560861A CN 106560861 A CN106560861 A CN 106560861A CN 201510635826 A CN201510635826 A CN 201510635826A CN 106560861 A CN106560861 A CN 106560861A
- Authority
- CN
- China
- Prior art keywords
- target object
- picture frame
- video
- angle point
- image
- 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
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/017—Detecting movement of traffic to be counted or controlled identifying vehicles
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
Abstract
The invention discloses an intelligent supervision method based on computer vision, wherein the method is used for processing a video monitoring image file which is obtained through a camera and stored in a computer. The method comprises the steps: employing a target tracking algorithm; tracking the target object through positive sequence and negative sequence reading; deciding a final moment of obtaining a clear snapshot image finally through the position of the target, thereby achieving the clear evidence obtaining in a larger range.
Description
Technical field
The invention belongs to place and road monitoring field, relate in particular to a kind of intelligent supervision method based on computer vision.
Background technology
Nowadays, obtained increasingly being widely applied based on the intelligent transportation monitoring and managing method of computer vision.There is the foundation of the image at the moment of act of violating regulations first as evidence obtaining in prior art, and be used to recognize with vehicle license plate in this image using vehicle, therefore the resolution ratio to camera has higher requirement, and limit the effective range of monitoring.In addition, prior art（A kind of supervising device and monitoring method of view-based access control model image, application number 201210127924, Authorization Notice No. 102693632B）By measure target velocity estimate its position carry out evidence obtaining for Car license recognition when, site error is larger, and reliability is not high.The present invention adopts a kind of new candid photograph evidence collecting method, under equal camera resolution, expands the scope of monitoring, and can guarantee that higher accuracy rate.
The present invention expands monitoring range, while having reached very high accuracy rate using the method for vision tracking.
The content of the invention
The present invention's aims at a kind of method for expanding monitoring range and improving monitoring accuracy rate of exploitation, and specifically, the present invention is employed the following technical solutions.
A kind of intelligent supervision method based on computer vision, obtains and preserves video monitoring image file in a computer, it is characterised in that the method comprising the steps of for processing by camera：
1）Video file is read in the normal order；
2）Detection sport foreground simultaneously filters out moving target object therein；
3）Whether detection moving target object there is object event；
4）If detecting moving target object object event occurs, current frame image is preserved, as first image p1 of evidence obtaining, and start video recording；
5）The above-mentioned moving target object for object event occur is tracked using target tracking algorism, records the positional information of moving target object, and the direction of motion of the moving target object is judged after n frames；
6）Subsequent video reading order is determined according to the direction of motion for being judged, even moving target object is towards camera position motion, then to read frame of video by positive sequence, if moving target object is remote from camera position motion, frame of video is read in reverse order；
7）Keep tracking the moving target object while frame of video is read, when the moving target object reaches predeterminated position, wherein the predeterminated position be in image target object close to camera and apparent position, current frame image is preserved as image forensic p2, stopping is recorded a video and preserves video file as evidence obtaining file.
Preferably, step 2）The screening of moving target object includes providing the size of the target object, and by size selection target object is gone out, and generates the external picture frame of target object.It is further preferred that the external picture frame of the target object of the generation is boundary rectangle frame or oval frame, such as when target object is vehicle, boundary rectangle frame can be generated, and in safety and protection monitoring field, such as when target object is human body, then can generate external oval frame.
In addition, in step 7）In, when stopping recording a video and preserving video recording, still further comprise from preservation video file and take the step of target object is the image forensic of same target in two reflection P1 and P2.
Preferably, there is the target object in than the target object in image forensic p2 closer to predeterminated position in this two images for being taken, it is more visible to ensure its feature, so as to when the part key character information of target object in image forensic p2 is imperfect, can be used in alternate image P2 or with the cross-referenced supplements of image P2.
Further, or even also include further by this two images for being taken in addition together with image forensic P1 and P2, this four images are synthesized a big image as evidence obtaining result, evidence convincingness is collected evidence and strengthened with easy-to-read.
Further, step 3）Whether detection moving target object object event occurs includes providing the demarcation information for obtaining the camera of this section of video monitoring image file, and the external picture frame of target object and demarcation information are compared to judge whether object event occur.
In addition, step 5）It is middle judge moving target object the method for the direction of motion be：According to the positional information of the n two field pictures for being recorded, the alternate position spike for calculating adjacent two frames moving target object obtains a vector, n-1 such vector is coupled together and obtains the direction of motion, judge to be remote from camera direction or close camera direction further according to demarcation information.
Further, judging the method for the direction of motion of moving target object includes：
5a）In a two field picture, for the target object having been detected by, its initial external picture frame R0 has been obtained, angle point has been extracted in the range of the picture frame, as initial angle point；
5b）Read next two field picture, in the two field picture, the picture frame with the nearest same shape in initial external picture frame R0 positions is found, and judge the distance of the picture frame and initial external picture frame R0, if wherein size of the distance of the picture frame and initial external picture frame R0 less than R0 itself, by the picture frame R1 is labeled as；
5c）To 5a）In angle point carry out optical flow tracking, in 5b）New angle point is obtained in the picture frame of middle reading, and the larger angle point of error is rejected, retain effective angle point, the minimum external picture frame of these effective angle points is labeled as into R3, if effectively angle point can not form external picture frame, angle point tracking failure, wherein, the method for rejecting angle point is：It is otherwise invalid angle point for effective angle point if angle point is in picture frame R0；
5d）If angle point is tracked successfully, but finds external picture frame failure, then R3 is assigned to into R1；If angle point tracking failure, and external picture frame success is found, then R1 is assigned to into R3.
Invention beneficial effect：1st, can monitor in a big way；2nd, higher accuracy rate can be reached；3rd, the method based on video recording is less demanding to computing power；The parameter requests such as the 4th, the well adapting to property of method based on tracking, and the resolution ratio to camera are relatively low；5th, roll lane line, the detection of guiding region and capture while realizing that downtown roads are ridden, solid line, Emergency Vehicle Lane are rolled on highway and is driven against traffic regulations and is stopped, and cart occupies the detection and candid photograph of the situation violating the regulations such as fast.
Description of the drawings
Fig. 1 is the schematic diagram of the video monitoring handling process based on computer vision of the present invention.
Specific embodiment
It is contemplated that the new method tracked using vision, expands monitoring range, detection is improved and the accuracy rate captured of collecting evidence, and while realize several functions.
Below by road vehicle supervision as a example by come to the present invention monitoring and managing method be described.
For the supervision of driving vehicle on road, process step is as follows：
1st, demarcated by demarcating module, completed initialization operation（After camera is installed, demarcate once, later in Same Scene, need not demarcate again in the case that camera is motionless）.
2nd, one section of video is enrolled at set intervals by camera, is preserved in a computer, optionally, it is also possible to intercepted a series of video image, and preserve in a computer, hereinafter all with video presentation.
3rd, computer constantly reads video file, and video file is processed, and has processed and delete the video after one section of video.Handling process is as shown in Figure 1.
Specifically, the flow process of computer disposal video file comprises the steps.
3.1st, video file is read according to normal sequence.
If the 3.2, reading last frame, EP (end of program) is otherwise continued executing with.
3.3rd, sport foreground is detected using background subtraction, and moving vehicle therein is gone out by size selection, removed pedestrian and other are disturbed.
3.4th, detect whether vehicle breaks rules and regulations, than such as whether line ball, if travel and stop in Emergency Vehicle Lane, whether cart occupies fast etc..Specifically detection method is：Solid line position is calibrated by step 1, and the position of Emergency Vehicle Lane and fast, the vehicle rectangle frame position obtained by step 3.3 is contrasted with demarcation information, judge whether to roll solid line, whether Emergency Vehicle Lane is located at, and whether fast is located at, while judging it is cart or dolly by rectangle frame size.
If the 3.5, being not detected by vehicles peccancy, step 3.1 is gone to, if detecting vehicles peccancy, preserve current frame image, as image forensic p1, and start avi file of recording a video.
3.6th, above-mentioned vehicles peccancy is tracked using target tracking algorism, records the positions and dimensions of each frame target vehicle, after n frames, judge direction of vehicle movement, concrete grammar is：The alternate position spike of adjacent two frames target vehicle is calculated, a vector can be obtained, n-1 such vector is coupled together, obtain the direction of motion, can obtain being remote from camera direction or close camera direction further according to demarcation information.
Specific tracking process can be further depicted as：（1）For a target vehicle having been detected by, its initial boundary rectangle frame is obtained, be designated as R0.First angle point is extracted in the range of the rectangle frame, as initial angle point.（2）Read next two field picture, the initial boundary rectangle frame contrast that the boundary rectangle frame and the two field picture that step 3.3 is detected is detected, obtain position and be separated by nearest rectangle frame, if the rectangle frame is with R0 at a distance of the size less than R0 itself, then the rectangle frame is labeled as into R1, otherwise, the failure of boundary rectangle frame is found.（3）It is right（1）In angle point carry out optical flow tracking,（2）New angle point is obtained in two field picture, and the larger angle point of error is rejected, retain effective angle point, the minimum enclosed rectangle of these effective angle points is labeled as into R3, if effectively angle point can not form boundary rectangle, angle point tracking fails, wherein, the method for rejecting angle point is：It is otherwise invalid angle point for effective angle point if angle point is in rectangle frame R0.（4）If angle point is tracked successfully, but finds the failure of boundary rectangle frame, then R3 is assigned to into R1；If angle point tracking failure, and boundary rectangle frame success is found, both failed for R3. if then R1 is assigned to into R3. tracking results, tracking failure.
If the 3.7, vehicle is travelled away from camera, backward reads frame of video and records present frame T simultaneously, if the close camera traveling of vehicle, positive sequence reads frame of video, and the vehicle is tracked.
3.8th, when above-mentioned vehicle reaches the position specified, current frame image is preserved, used as image forensic p2, the position is the position close to camera, the image for obtaining can see clearly information of vehicles, such as the clear discernible position of car plate of target vehicle during electronic police application.And stop video recording, preserve the video recording avi file, as evidence obtaining file.If being now backward reading, reading manner is changed to into positive sequence, and starts to read from P+n+1 frames, go to step 3.1.
The method of the present invention judges direction of vehicle movement by target tracking algorism, the final moment for obtaining clear candid photograph image is determined by the position of target, by positive sequence and the tracking target vehicle of backward, the interior clear evidence obtaining to target vehicle in a big way can be realized, can understand that monitoring in time is rolled solid line, Emergency Vehicle Lane and driven against traffic regulations and stop etc. act of violating regulations, and the detection and candid photograph for realizing situations such as cart occupies fast.
Embodiment
Below the present invention is described in detail with instantiation.
The basic hardware composition of the present invention includes video camera and supervision processor（Computer）.
Video camera is any device that can obtain realtime graphic, including gunlock, high-speed ball-forming machine etc..During installation, camera review visual field it is long and narrow square consistent with road direction, such as, resolution of video camera is 1920X1080, then obtain image in 1920 corresponding sides it is consistent with road direction.Can so cause monitoring range wider.Resolution of video camera require for meet vehicle view field image most nearby when can see car plate.
As a example by roll yellow line detection, when installing video camera, the long side of view field image is consistent with road direction, that is, consistent with yellow line direction, adjusts camera angle so that field coverage is as long as possible, it is only necessary to which meeting nearby vehicle can see car plate clearly.After video camera installation, demarcated by demarcating module, i.e., the relation set up between image coordinate system and actual coordinates using camera parameters.
Camera acquisition view data, and supervision processor is passed to by modes such as wire communication modes, supervision processor is processed the image for collecting.
Next, supervision processor control camera records at set intervals one section of video, and store.Supervise processor simultaneously in order to process each section of video, as a example by rolling yellow line and detect, flow process is as follows：
1st, video file is read according to normal sequence；
If the 2, reading last frame, EP (end of program) is otherwise continued executing with；
3rd, updated using average background and obtain real-time background image, and sport foreground is obtained by background subtraction.Further more complete foreground image is obtained by morphological operation and average value filtering, filter noise.After obtaining more complete sport foreground, the boundary rectangle frame of moving target is obtained using contour detecting, with reference to the information of demarcation, the actual size of each rectangle frame can be obtained, so as to filter the rectangle frame for being unsatisfactory for vehicle size.Finally obtain the boundary rectangle frame of all vehicles；
4th, by whether judge above-mentioned boundary rectangle frame and calibrate come yellow line intersect judging vehicle whether line ball, intersect i.e. line ball, it is non-intersect then without line ball；
If the 5, being not detected by line ball vehicle, step 1 is gone to, if detecting line ball vehicle, preserve current frame image, as first image p1 of evidence obtaining, and start avi file of recording a video；
6th, above-mentioned vehicles peccancy is tracked using target tracking algorism, records each frame target vehicle boundary rectangle frame, this has just obtained the size and positional information of vehicle, after n frames, judges direction of vehicle movement, and concrete grammar is：The alternate position spike of adjacent two frames target vehicle is calculated, a vector can be obtained, n-1 such vector is coupled together, obtain the direction of motion, can obtain being remote from camera direction or close camera direction further according to demarcation information.In the image that such as video camera is obtained, road direction is X-direction, and knowable to the information of demarcation, along X-axis augment direction（Namely to the right）It is, away from camera direction, to reduce along X-axis（To the left）To be close to camera direction.By the X-coordinate information of vehicle location in n frames, the x-axis direction vector of adjacent two frame is obtained, be away from camera direction, if to the left, to be close to camera direction if to the right by n-1 vector addition；
If the 7, vehicle is travelled away from camera, backward reads frame of video and records present frame T simultaneously, if vehicle is travelled near camera, positive sequence reads frame of video, and the vehicle is tracked；
8th, when above-mentioned vehicle reaches the position specified, current frame image is preserved, used as image forensic p2, the position is the position close to camera, and the image for obtaining can see clearly information of vehicles.Wherein the position be in video target object close to camera and apparent position, such as the clear discernible position of car plate of target vehicle during electronic police application.Then current frame image is preserved as image forensic p2, stopping is recorded a video and preserves video recording avi file as evidence obtaining file, can be with from preserving the image forensic for taking that target object in two reflections P1 and P2 is same target in video file, and this two images have one to need to consider target object closer to predeterminated position, it is more visible to ensure its feature, when target object key character information is imperfect in image forensic p2, for alternate image P2, and four images are synthesized a big image as evidence obtaining result.
If being now backward reading, reading manner is changed to into positive sequence, and starts to read from P+n+1 frames, go to step 1.
Although above description is illustrated so that road vehicle is supervised in violation of rules and regulations as an example, it is to be appreciated that the present invention is applicable not only to road monitoring; can be applicable to other occasions; such as enterprise, residential quarter security protection, hazardous area monitoring etc., specific protection domain should be defined by claims.
For example in safety-security area, particular event can occur as detection evidence obtaining event using special exercise target object, such as using human body proximity parking lot vehicle as treating evidence obtaining event.Now, the external picture frame of human body can be set as ellipse, take human body feature point as the test point of external profile, then detect event to be measured by the relevant position of the external profile and the parking space information demarcated.When needing, evidence obtaining is analyzed to event using the inventive method.
Embodiments of the present invention are described in detail above in conjunction with the drawings and specific embodiments, but the invention is not restricted to above-mentioned embodiment, in the ken that art those of ordinary skill possesses, can be making a variety of changes on the premise of without departing from present inventive concept.
Claims (9)
1. a kind of intelligent supervision method based on computer vision, obtains and preserves video monitoring image file in a computer, it is characterised in that the method comprising the steps of for processing by camera：
1）Video file is read in the normal order；
2）Detection sport foreground simultaneously filters out moving target object therein；
3）Whether detection moving target object there is object event；
4）If detecting moving target object object event occurs, current frame image is preserved, as first image p1 of evidence obtaining, and start video recording；
5）The above-mentioned moving target object for object event occur is tracked using target tracking algorism, records the positional information of moving target object, and the direction of motion of the moving target object is judged after n frames；
6）Subsequent video reading order is determined according to the direction of motion for being judged, even moving target object is proximate to camera position motion, then reads frame of video by positive sequence, if moving target object is remote from camera position motion, frame of video is read in reverse order；
7）Keep tracking the moving target object while frame of video is read, when the moving target object reaches predeterminated position, wherein the predeterminated position be in video target object close to camera and apparent position, current frame image is preserved as image forensic p2, stopping is recorded a video and preserves video file as evidence obtaining file.
2. the intelligent supervision method of computer vision is based on as claimed in claim 1, it is characterised in that step 2）The screening of moving target object includes providing the size of the target object, and by size selection target object is gone out, and generates the external picture frame of target object.
3. the intelligent supervision method of computer vision is based on as claimed in claim 2, it is characterised in that the external picture frame of the target object for being generated is boundary rectangle picture frame or external oval picture frame.
4. the intelligent supervision method of computer vision is based on as claimed in claim 1, it is characterised in that in step 7）In, when stopping recording a video and preserving video recording, the image forensic for taking that target object in two reflection P1 and P2 is same target is still further comprised from preservation video file.
5. the intelligent supervision method of computer vision is based on as claimed in claim 4, it is characterized in that, there is the target object in than the target object in image forensic p2 closer to predeterminated position in this two images for being taken, it is more visible to ensure its feature, so as to when the part key character information of target object in image forensic p2 is imperfect, can be used in alternate image P2 or with the cross-referenced supplements of image P2.
6. the intelligent supervision method based on computer vision as described in claim 4 or 5, it is characterized in that, also include this two images that will be taken in addition together with image forensic P1 and P2, this four images are synthesized a big image as evidence obtaining result, evidence convincingness is collected evidence and strengthened with easy-to-read.
7. the intelligent supervision method of computer vision is based on as claimed in claim 1, it is characterised in that step 3）Whether detection moving target object object event occurs includes providing the demarcation information for obtaining the camera of this section of video monitoring image file, and the external picture frame of target object and demarcation information are compared to judge whether object event occur.
8. the intelligent supervision method of computer vision is based on as claimed in claim 1, it is characterised in that step 5）It is middle judge moving target object the method for the direction of motion be：According to the positional information of the n two field pictures for being recorded, the alternate position spike for calculating adjacent two frames moving target object obtains a vector, n-1 such vector is coupled together and obtains the direction of motion, judge to be remote from camera direction or close camera direction further according to demarcation information.
9. the intelligent supervision method of computer vision is based on as claimed in claim 8, it is characterised in that judging the method for the direction of motion of moving target object includes：
5a）In a two field picture, for the target object having been detected by, its initial external picture frame R0 has been obtained, angle point has been extracted in the range of the picture frame, as initial angle point；
5b）Read next two field picture, in the two field picture, find the external picture frame with the nearest same shape in initial external picture frame R0 positions, and judge the external picture frame of the same shape and the distance of initial external picture frame R0, if wherein size of the distance of the picture frame and initial external picture frame R0 less than R0 itself, by the picture frame R1 is labeled as；
5c）To 5a）In angle point carry out optical flow tracking, in 5b）New angle point is obtained in the picture frame of middle reading, and the larger angle point of error is rejected, retain effective angle point, the minimum external picture frame of these effective angle points is labeled as into R3, if effectively angle point can not form external picture frame, angle point tracking failure, wherein, the method for rejecting angle point is：It is otherwise invalid angle point for effective angle point if angle point is in picture frame R0；
5d）If angle point is tracked successfully, but finds external picture frame failure, then R3 is assigned to into R1；If angle point tracking failure, and external picture frame success is found, then R1 is assigned to into R3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510635826.XA CN106560861A (en) | 2015-09-30 | 2015-09-30 | Intelligent supervision method based on computer vision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510635826.XA CN106560861A (en) | 2015-09-30 | 2015-09-30 | Intelligent supervision method based on computer vision |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106560861A true CN106560861A (en) | 2017-04-12 |
Family
ID=58485132
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510635826.XA Pending CN106560861A (en) | 2015-09-30 | 2015-09-30 | Intelligent supervision method based on computer vision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106560861A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107844772A (en) * | 2017-11-09 | 2018-03-27 | 汕头职业技术学院 | A kind of motor vehicle automatic testing method based on movable object tracking |
| CN108229475A (en) * | 2018-01-03 | 2018-06-29 | 深圳中兴网信科技有限公司 | Wireless vehicle tracking, system, computer equipment and readable storage medium storing program for executing |
| CN109117702A (en) * | 2018-06-12 | 2019-01-01 | 深圳中兴网信科技有限公司 | The detection and count tracking method and system of target vehicle |
| CN109615858A (en) * | 2018-12-21 | 2019-04-12 | 深圳信路通智能技术有限公司 | A kind of intelligent parking behavior judgment method based on deep learning |
| CN109753940A (en) * | 2019-01-11 | 2019-05-14 | 京东方科技集团股份有限公司 | Image processing method and device |
| CN110033476A (en) * | 2018-01-11 | 2019-07-19 | 富士通株式会社 | Target velocity estimation method, device and image processing equipment |
| CN110728843A (en) * | 2019-09-10 | 2020-01-24 | 浙江大华技术股份有限公司 | Vehicle snapshot method, vehicle snapshot device, and storage medium |
| CN110991397A (en) * | 2019-12-17 | 2020-04-10 | 深圳市捷顺科技实业股份有限公司 | Traveling direction determining method and related equipment |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101216996A (en) * | 2007-12-28 | 2008-07-09 | 云南澜博科技有限公司 | A method for monitoring vehicle offences by video |
| CN101593422A (en) * | 2009-06-15 | 2009-12-02 | 无锡骏聿科技有限公司 | A kind of traffic violation detection method based on video frequency tracking and pattern-recognition |
| CN101789177A (en) * | 2010-02-05 | 2010-07-28 | 南京航空航天大学 | Device and method for detecting and tracking vehicles crossing and pressing the yellow line and for capturing vehicle information |
| CN101937614A (en) * | 2010-06-12 | 2011-01-05 | 北京中科卓视科技有限责任公司 | Plug and play comprehensive traffic detection system |
| CN102073851A (en) * | 2011-01-13 | 2011-05-25 | 北京科技大学 | Method and system for automatically identifying urban traffic accident |
| CN102693632A (en) * | 2012-04-27 | 2012-09-26 | 徐贵力 | Monitoring device and monitoring method based on visual image |
| CN103065470A (en) * | 2012-12-18 | 2013-04-24 | 浙江工业大学 | Detection device for behaviors of running red light of vehicle based on machine vision with single eye and multiple detection faces |
| CN103325258A (en) * | 2013-06-24 | 2013-09-25 | 武汉烽火众智数字技术有限责任公司 | Red light running detecting device and method based on video processing |
-
2015
- 2015-09-30 CN CN201510635826.XA patent/CN106560861A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101216996A (en) * | 2007-12-28 | 2008-07-09 | 云南澜博科技有限公司 | A method for monitoring vehicle offences by video |
| CN101593422A (en) * | 2009-06-15 | 2009-12-02 | 无锡骏聿科技有限公司 | A kind of traffic violation detection method based on video frequency tracking and pattern-recognition |
| CN101789177A (en) * | 2010-02-05 | 2010-07-28 | 南京航空航天大学 | Device and method for detecting and tracking vehicles crossing and pressing the yellow line and for capturing vehicle information |
| CN101937614A (en) * | 2010-06-12 | 2011-01-05 | 北京中科卓视科技有限责任公司 | Plug and play comprehensive traffic detection system |
| CN102073851A (en) * | 2011-01-13 | 2011-05-25 | 北京科技大学 | Method and system for automatically identifying urban traffic accident |
| CN102693632A (en) * | 2012-04-27 | 2012-09-26 | 徐贵力 | Monitoring device and monitoring method based on visual image |
| CN103065470A (en) * | 2012-12-18 | 2013-04-24 | 浙江工业大学 | Detection device for behaviors of running red light of vehicle based on machine vision with single eye and multiple detection faces |
| CN103325258A (en) * | 2013-06-24 | 2013-09-25 | 武汉烽火众智数字技术有限责任公司 | Red light running detecting device and method based on video processing |
Non-Patent Citations (1)
| Title |
|---|
| 张洪斌等: "《面向城市路口的高清晰智能监控系统研究》", 《四川大学学报（工程科学版）》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107844772A (en) * | 2017-11-09 | 2018-03-27 | 汕头职业技术学院 | A kind of motor vehicle automatic testing method based on movable object tracking |
| CN108229475B (en) * | 2018-01-03 | 2021-08-31 | 深圳中兴网信科技有限公司 | Vehicle tracking method, system, computer device and readable storage medium |
| CN108229475A (en) * | 2018-01-03 | 2018-06-29 | 深圳中兴网信科技有限公司 | Wireless vehicle tracking, system, computer equipment and readable storage medium storing program for executing |
| CN110033476A (en) * | 2018-01-11 | 2019-07-19 | 富士通株式会社 | Target velocity estimation method, device and image processing equipment |
| CN109117702A (en) * | 2018-06-12 | 2019-01-01 | 深圳中兴网信科技有限公司 | The detection and count tracking method and system of target vehicle |
| CN109117702B (en) * | 2018-06-12 | 2022-01-25 | 深圳中兴网信科技有限公司 | Target vehicle detection, tracking and counting method and system |
| CN109615858A (en) * | 2018-12-21 | 2019-04-12 | 深圳信路通智能技术有限公司 | A kind of intelligent parking behavior judgment method based on deep learning |
| CN109753940A (en) * | 2019-01-11 | 2019-05-14 | 京东方科技集团股份有限公司 | Image processing method and device |
| CN109753940B (en) * | 2019-01-11 | 2022-02-22 | 京东方科技集团股份有限公司 | Image processing method and device |
| CN110728843B (en) * | 2019-09-10 | 2021-08-31 | 浙江大华技术股份有限公司 | Vehicle snapshot method, vehicle snapshot device, and storage medium |
| CN110728843A (en) * | 2019-09-10 | 2020-01-24 | 浙江大华技术股份有限公司 | Vehicle snapshot method, vehicle snapshot device, and storage medium |
| CN110991397A (en) * | 2019-12-17 | 2020-04-10 | 深圳市捷顺科技实业股份有限公司 | Traveling direction determining method and related equipment |
| CN110991397B (en) * | 2019-12-17 | 2023-08-04 | 深圳市捷顺科技实业股份有限公司 | Travel direction determining method and related equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106560861A (en) | Intelligent supervision method based on computer vision | |
| US9224049B2 (en) | Detection of static object on thoroughfare crossings | |
| EP0567059B1 (en) | Object recognition system using image processing | |
| CN103366569B (en) | The method and system of real-time grasp shoot traffic violation vehicle | |
| US7262710B2 (en) | Collision time estimation apparatus for vehicles, collision time estimation method for vehicles, collision alarm apparatus for vehicles, and collision alarm method for vehicles | |
| CN108091142A (en) | For vehicle illegal activities Tracking Recognition under highway large scene and the method captured automatically | |
| CN101964145A (en) | Automatic license plate recognition method and system | |
| CN105809975A (en) | Abnormal parking judgment method and abnormal parking judgment device | |
| CN110298300B (en) | Method for detecting vehicle illegal line pressing | |
| CN110738150B (en) | Camera linkage snapshot method and device and computer storage medium | |
| CN105679043A (en) | 3D radar intelligent bayonet system and processing method thereof | |
| CN112367475B (en) | Traffic incident detection method and system and electronic equipment | |
| CN105818743A (en) | Anticollision method and system used when vehicle doors are opened | |
| CN106558224A (en) | A kind of traffic intelligent monitoring and managing method based on computer vision | |
| CN111292530A (en) | Method, device, server and storage medium for processing violation pictures | |
| Choi et al. | Cut-in vehicle warning system exploiting multiple rotational images of SVM cameras | |
| CN101175203A (en) | Real-time video target tracking method | |
| Yang | Estimation of vehicle's lateral position via the Lucas-Kanade optical flow method | |
| CN112489125A (en) | Automatic detection method and device for storage yard pedestrians | |
| CN110782485A (en) | Vehicle lane change detection method and device | |
| CN112906428A (en) | Image detection area acquisition method and space use condition judgment method | |
| CN116152753A (en) | Vehicle information identification method and system, storage medium and electronic device | |
| KR102434154B1 (en) | Method for tracking multi target in traffic image-monitoring-system | |
| CN114973157A (en) | Vehicle separation method, electronic device, and computer-readable storage medium | |
| CN114399537B (en) | Vehicle tracking method and system for target personnel |
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: 20170412 |