CN104809437A - Real-time video based vehicle detecting and tracking method - Google Patents
Real-time video based vehicle detecting and tracking method Download PDFInfo
- Publication number
- CN104809437A CN104809437A CN201510210294.5A CN201510210294A CN104809437A CN 104809437 A CN104809437 A CN 104809437A CN 201510210294 A CN201510210294 A CN 201510210294A CN 104809437 A CN104809437 A CN 104809437A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- result
- real
- target
- tracking
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a real-time video based vehicle detecting and tracking method. The method includes S101, acquiring an initial sample set; S102, training an initial classifier; S103, iterating a training classifier; S104, generating a vehicle candidate area; S105, confirming and tracking vehicles; S106, counting traffic flowrate; S107, performing algorithm interaction. The real-time video based vehicle detecting and tracking method has the advantages that the vehicles appearing in a real-time video can be detected and tracked stably, the traffic flowrate can be counted, high robustness, adaptability to noise, illumination and weather changes and a certain adaptability to shielding are achieved, and the method is high in processing speed and accuracy and capable of meeting operation requirements of a real-time system.
Description
Technical field
The present invention relates to moving vehicles detection and tracking technical field, particularly relate to a kind of moving vehicles detection and tracking method based on real-time video.
Background technology
Moving vehicles detection and tracking refers to from the information of vehicles by occurring analysis sequence of video images automatic acquisition picture, and carries out tenacious tracking to same vehicle target in continuous videos.Moving vehicles detection and tracking technology is the gordian technique in the field such as intelligent transportation system, intelligent security guard, can well assist the work of related personnel and increase work efficiency, and therefore becomes the study hotspot of digital image processing field.Common moving vehicles detection and tracking has road vehicle detection and tracking, gateway moving vehicles detection and tracking, the moving vehicles detection and tracking etc. of satellite vertical view.
Moving vehicles detection and tracking based on video has higher practical value and is widely used, and a lot of scholar is devoted to study related algorithm in recent years.Emerge a lot of vehicle checking method up to now, the people such as doctor Zehang (IEEE member) in Nevada ,Usa state university's Reynolds branch school are doing large quantity research and summary to existing method, and on " IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINEINTELLIGENCE ", published thesis " On-Road Vehicle Detection:A Review " in 2006, in literary composition, existing method is roughly summarised as and generates candidate region and candidate region and verify two steps, namely quick position may be the region of vehicle in the picture, then confirm whether it is vehicle further.Improved vehicle detecting algorithm although researchist is also continuous in the last few years, Integral Thought still follows this two steps.The method of the generation of candidate region mainly contains three classes, is based on priori, based on stereoscopic vision and based drive method respectively.Method based on priori obtains vehicle candidate region by features such as the symmetry in analysis chart picture, color, shade, angle point, edge, texture, car lights.Analyze the stereoscopic features of vehicle mainly through disparity map, inverse perspective mapping etc. that two cameras obtain image based on the method for stereoscopic vision.Based drive method mainly through the methods such as optical flow method, motion vector method, frame difference method and background subtraction method obtain moving object information thus as vehicle candidate.The method of candidate region checking is mainly divided into based on template and two class methods based on outward appearance.Method based on template is confirmed compared with the template image obtained in advance by the image of candidate region, and these class methods are comparatively large by the impact of template, can not be correctly detected when candidate's vehicle and template differ greatly.Method based on outward appearance mainly utilizes the method study vehicle characteristics of machine learning, and the feature of candidate region is confirmed compared with vehicle characteristics, this kind of methods and results is more stable, and in most of the cases performance is good, but testing result is larger by the impact of training sample set.The method for tracking target of current main flow has kalman filter method, and Meanshift method and Camshift method etc., these methods are monotrack method, and the method for multiple target tracking is mostly the improvement based on these methods.It should be noted that these trackings need first to specify tracking target, then follow the tracks of in the video sequence, generally only have to carry out vehicle detection at the beginning when combining with vehicle checking method, just stop after target being detected detecting and starting to follow the tracks of this target, the benefit done like this is that time overhead is few, and can suppress error-detecting.But this mode also has its very important drawback, and error detection result will be directly passed to tracker, and has no chance to be corrected, if testing result is bad, the reliability of tracker directly can be injured.
The leading indicator weighing vehicle detecting algorithm performance has robustness, accuracy and processing speed.But existing method is difficult to take into account this three indexs simultaneously, although its robustness of such as frame difference method and edge detection method is higher, processing speed is also very fast, and the accuracy of these two kinds of methods is lower; Although and optical flow method and the accuracy of motion vector method still can, the poor and processing speed of robustness is difficult to, and cannot meet the application demand of real-time system; Background subtraction method is because have higher accuracy and processing speed faster, and application is comparatively extensive, but due to noise, illumination and weather sensitivity, its robustness is also a problem that can not be ignored.
Summary of the invention
The object of the invention is to, by a kind of moving vehicles detection and tracking method based on real-time video, solve the problem that above background technology part is mentioned.
For reaching this object, the present invention by the following technical solutions:
Based on a moving vehicles detection and tracking method for real-time video, it comprises the steps:
S101, obtain initial sample set:
Image results in file system is classified, is positioned over respectively in different files, obtain initial samples pictures collection; Adopt autoexec to obtain the absolute path list of samples pictures in each file respectively, and give different marks to the file in each file, finally the sample absolute path list of tape label is incorporated in a text;
S102, training preliminary classification device:
The initial sample set training classifier utilizing step S101 to obtain, reads in each sample and mark thereof successively, analyzes the proper vector that its feature obtains each sample, carries out Supervised classification to proper vector, obtain preliminary classification device;
S103, repetitive exercise sorter:
The preliminary classification device utilizing step S102 to obtain is classified to the moving object in video sequence, and the local area image at moving object place is added timestamp name be stored in file system by classification results mark, if user is satisfied to classification results, without the need to carrying out repetitive exercise, if dissatisfied, find out unsatisfied misclassification result, and be added in the sample set file of correct class, method described in step S101 is utilized to obtain new training sample set text, then method described in step S102 is utilized to carry out repetitive exercise, so repeatedly until user is to the satisfied repetitive exercise then completing sorter of classification results, sorter writing in files is saved in file system,
S104, vehicle candidate region generate:
Read in sorter file, detect the moving region image occurred in video sequence successively, the feature analyzing moving region image obtains its proper vector, proper vector is sent into sorter to carry out classifying and obtaining classification results, if result is vehicle, thinks that this moving region image comprises vehicle, complete the generation of vehicle candidate region with this;
S105, vehicle confirm and follow the tracks of:
The vehicle detection result in every two field picture is obtained in step S104, first be benchmark with present frame in real-time video, to all vehicle detection results in each vehicle detection result traversal previous frame of present frame, similarity indices will be met and minimum one of similarity measurement result is considered as the matching result of current goal; Same, take previous frame as benchmark, to the result of each testing result traversal present frame of previous frame, will similarity indices be met and minimum one of similarity measurement result is considered as the matching result of previous frame target; Finally, if previous frame exists and present frame lost target is considered as target disappearance, if the target that present frame occurs does not appear in previous frame, be considered as fresh target and distribute new mark for it, if two targets of present frame and previous frame match each other, be considered as same target and follow the tracks of with same tag.
Especially, the described moving vehicles detection and tracking method based on real-time video also comprises:
S106, a wagon flow statistics:
Measurement type is divided into three kinds: the vehicle fleet occurred in picture, sails vehicle fleet into, outgoing vehicles sum; The statistics of vehicle fleet is carried out according to the result of vehicle tracking in step S105, only has and just counts when fresh target occurs; Sail into and add up according to the mode of dummy line with outgoing vehicles, be in line, then think that it is for sailing vehicle into when original state is vehicle's current condition outside line, carrying out counting and its original state being changed is set in line, and wherein, the statistical method of outgoing vehicles is identical with the statistical method of sailing vehicle into.
Especially, the described moving vehicles detection and tracking method based on real-time video also comprises:
S107, algorithm are mutual:
Because the local area image of the vehicle target in each two field picture all can be stored in file system, this testing result is comprised the information write server database in camera number, writing time, vehicle target image results path simultaneously; When reaching the vehicle flowrate time interval, by present period data write into Databasce and by counter O reset at every turn; Interactive program can the state in real-time monitoring data storehouse, carries out playing up display result according to data inserting when there being data to insert; Whenever reaching running state of programs daily record interval, by in program current state write journal file, to have good grounds when breaking down, on the other hand, program provides configuration file to revise parameter for user, and user can revise the supplemental characteristic comprising the vehicle flowrate time interval, running state of programs daily record interval, frame picture process dutycycle;
Especially, described step S105 also comprises: utilize vehicle tracking result to improve vehicle detection effect, the target of tenacious tracking preset times just confirms as vehicle, in this, as vehicle detection result, preserve the image results of vehicle place regional area to file system and by testing result data data inserting storehouse simultaneously.
Especially, described step S105 comprises further: for the result of tenacious tracking, if suddenly disappeared, gives predictive compensation and presets frame number, if again occurred with internal object at default frame number, continues to follow the tracks of, otherwise thinks track rejection.
The moving vehicles detection and tracking method tool based on real-time video that the present invention proposes has the following advantages: one, robustness is high: utilize tracking results effectively to overcome the impact of noise to the correction that testing result is carried out, the repetitive exercise of sorter effectively overcomes the impact of the change such as illumination and weather, and can obtain better testing result by exptended sample collection at any time.Two, accuracy is high: such as through 4 repetitive exercise, when sample set quantity is 4146, the size of sorter is 89.1M, 3415 mistake point results are only had in 62448 classification results that continuous testing results obtains, accuracy reaches 94.53%, and continuation iteration can reach higher accuracy.Three, processing speed is fast, and program scale is controlled: all video input resolution adjustments are that 640*360 processes again by such as program, and therefore picture quality changes does not affect processing speed, and average every frame processing speed is 20ms.Be 4M in video code flow bandwidth, resolution is 1280*720, and frame per second is 25, and when frame process dutycycle is 5ms, CPU usage is about 45%, and memory usage is about 142M.And when frame process dutycycle becomes other parameter constants of 200ms, CPU usage is reduced to about 24%, memory usage is about 137M.When optimum configurations is bandwidth 512K, resolution 640*360, frame per second 15, during dutycycle 200ms, CPU usage is about 17%, and memory usage is about 115M, can smoothly in low configuration computing machine run completely.Four, there is good interactivity: program performs different functional modules according to difference input instruction, operational factor can be changed by user, variation Output rusults facilitates Front End to play up display effect, the mode being stored to database makes this program can be completely independent with Front End, and front end even can adopt different programming languages.
Accompanying drawing explanation
The moving vehicles detection and tracking method flow diagram based on real-time video that Fig. 1 provides for the embodiment of the present invention;
The vehicle candidate region that Fig. 2 provides for the embodiment of the present invention generates method flow diagram;
The vehicle that Fig. 3 provides for the embodiment of the present invention confirms and tracking process flow diagram;
The vehicle tracking process schematic that Fig. 4 provides for the embodiment of the present invention.
Embodiment
Below in conjunction with drawings and Examples, the invention will be further described.Be understandable that, specific embodiment described herein is only for explaining the present invention, but not limitation of the invention.It also should be noted that, for convenience of description, illustrate only part related to the present invention in accompanying drawing but not full content, unless otherwise defined, all technology used herein and scientific terminology are identical with belonging to the implication that those skilled in the art of the present invention understand usually.Term used herein, just in order to describe specific embodiment, is not intended to be restriction the present invention.
Please refer to shown in Fig. 1, the moving vehicles detection and tracking method flow diagram based on real-time video that Fig. 1 provides for the embodiment of the present invention.
Moving vehicles detection and tracking method based on real-time video in the present embodiment specifically comprises the steps:
S101, obtain initial sample set:
The all moving objects occurred in capture video sequences, and the image of moving object place regional area is stored in file system, by user, manual sort is carried out to the image results in file system, be positioned over respectively in different files, obtain initial samples pictures collection.As being classified as follows: 0 non-vehicle class, the positive and negative Noodles of 1 vehicle, 2 vehicle at night classes, 3 vehicle side Noodles, 4 local vehicle classes, 5 have shielding automobile class.Adopt autoexec to obtain the absolute path list of samples pictures in each file respectively, and give corresponding figure notation to the file in each file, finally the sample absolute path list of these tape labels is incorporated in same text.
S102, training preliminary classification device:
Training preliminary classification device.The initial sample set training classifier utilizing step S101 to obtain, read in each sample and mark thereof successively, such as can first by unified for the image size being adjusted to 64*64, then 1764 dimensional feature vectors that its histograms of oriented gradients obtains each sample are analyzed, finally adopt support vector machine to carry out Supervised classification to these proper vectors, obtain preliminary classification device.
S103, repetitive exercise sorter:
The preliminary classification device utilizing step S102 to obtain is classified to the moving object in video sequence, and the local area image at moving object place is added timestamp name be stored in file system by classification results mark, if user is satisfied to classification results, without the need to carrying out repetitive exercise, if dissatisfied, need to find out unsatisfied misclassification result, and be added in the sample set file of correct class, method described in step S101 is utilized to obtain new training sample set text, then method described in step S102 is utilized to carry out repetitive exercise, so repeatedly until user is to the satisfied repetitive exercise then completing sorter of classification results.Sorter writing in files is saved in file system.
S104, vehicle candidate region generate:
First sorter file is read in, detect the moving region image occurred in video sequence successively, the feature analyzing moving region image obtains its proper vector, proper vector is sent into sorter to carry out classifying and obtaining classification results, if result is vehicle, thinks that this moving region image comprises vehicle, complete the generation of vehicle candidate region with this.Specifically as shown in Figure 2: initiation parameter; Be loaded into sorter; Obtain next frame image; Judge whether image is empty, if it is empty, then terminates, otherwise current results is copied to previous frame; Detect moving region; Sorter is utilized to obtain moving region key words sorting; Determine whether vehicle, if not, then obtain next frame image, if so, then preserve the information of candidate region; Vehicle confirms and follows the tracks of; Judge whether to exit, if so, then terminate, if not, then obtain next frame image.
S105, vehicle confirm and follow the tracks of:
As shown in Figure 3 and Figure 4, specific as follows: in step S104, to obtain the vehicle candidate region result in every two field picture, first be benchmark with present frame in real-time video, each vehicle detection result Ci of present frame is traveled through to all vehicle detection results in previous frame, will similarity indices be met and the minimum Pj of similarity measurement result is considered as matching result and the Ci->Pj of current goal.Similar, take previous frame as benchmark, each testing result Pi of previous frame is traveled through to the result of present frame, will similarity indices be met and the minimum Cj of similarity measurement result is considered as matching result and the Pi->Cj of previous frame target.If last previous frame exists and present frame lost target is considered as target disappearance, if the target that present frame occurs does not appear in previous frame, be considered as fresh target and distribute new mark for it, if two targets of present frame and previous frame match each other, namely Ci->Pj is met and Pj->Ci, then Ci and Pj is considered as same target, and follows the tracks of with the mark identical with Pj Ci.In Fig. 4, C1, C2, C3 of present frame is respectively the tracking results of P1, P2, P3 of previous frame, and target P 4 disappears, and C4 is fresh target.Utilize vehicle tracking result to improve vehicle detection effect simultaneously, the target that tenacious tracking is more than m time just confirms as vehicle, in this, as vehicle detection result, preserve the image results of vehicle place regional area to file system and by testing result data data inserting storehouse simultaneously, effectively eliminate partial noise interference; If suddenly disappeared for the result of tenacious tracking, give predictive compensation n frame, if n frame occurs again with internal object, continue to follow the tracks of, otherwise think track rejection, efficiently solve and detect unstable problem of interrupting.
S106, a wagon flow statistics:
Measurement type is divided into three kinds, is the vehicle fleet occurred in picture in timing statistics respectively, sails vehicle fleet and outgoing vehicles sum into.The statistics of vehicle fleet is carried out according to the result of vehicle tracking in step S105, only has and just count when fresh target occurs, revise the count flag position of this target after counting, and the target that each frame occurs afterwards is all inherited this zone bit and prevented repeat count.Sailing into outgoing vehicles is that mode according to dummy line is added up, and in order to prevent the fluctuation of vehicle detection result coordinate, dummy line arranges certain width.Relative position according to itself and dummy line when target occurs first arranges its original state, and when the online outer vehicle's current condition of original state is then think in line that it sails into, carry out counting and its original state changed and be set in line, outgoing vehicles in like manner.
S107, algorithm are mutual:
As mentioned above, the local area image of the vehicle target in each two field picture all can be stored in file system, simultaneously by information write server databases such as the camera number of this testing result, writing time, vehicle target image results paths.When reaching the time interval of vehicle flowrate, by present period data write into Databasce and by counter O reset at every turn.Interactive program can the state in real-time monitoring data storehouse, carries out playing up display result according to data inserting when there being data to insert.Whenever reaching running state of programs daily record interval, by program current state write journal file, have good grounds when guaranteeing to break down with this.On the other hand, program provides configuration file to revise parameter for user, and user can revise the parameters such as the vehicle flowrate time interval, running state of programs daily record interval, frame picture process dutycycle, video camera ID, video camera and bayonet socket relative position, dummy line coordinate, image results store path, Database Properties, video address.Amendment frame picture process dutycycle can control program scale, and effectively control CPU usage and the memory usage of this program, such as, when under default parameters state, frame process dutycycle is 5ms, CPU usage is about 45%, and memory usage is about 142M.And when frame process dutycycle becomes other parameter constants of 200ms, CPU usage is reduced to about 24%, memory usage is about 137M, guarantees can both reach optimal operational condition in the computing machine of difference configuration with this.
Technical scheme advantage of the present invention is as follows: one, robustness is high: utilize tracking results effectively to overcome the impact of noise to the correction that testing result is carried out, the repetitive exercise of sorter effectively overcomes the impact of the change such as illumination and weather, and can obtain better testing result by exptended sample collection at any time.Two, accuracy is high: such as through 4 repetitive exercise, when sample set quantity is 4146, the size of sorter is 89.1M, 3415 mistake point results are only had in 62448 classification results that continuous testing results obtains, accuracy reaches 94.53%, and continuation iteration can reach higher accuracy.Three, processing speed is fast, and program scale is controlled: all video input resolution adjustments are that 640*360 processes again by such as program, and therefore picture quality changes does not affect processing speed, and average every frame processing speed is 20ms.Be 4M in video code flow bandwidth, resolution is 1280*720, and frame per second is 25, and when frame process dutycycle is 5ms, CPU usage is about 45%, and memory usage is about 142M.And when frame process dutycycle becomes other parameter constants of 200ms, CPU usage is reduced to about 24%, memory usage is about 137M.When optimum configurations is bandwidth 512K, resolution 640*360, frame per second 15, during dutycycle 200ms, CPU usage is about 17%, and memory usage is about 115M, can smoothly in low configuration computing machine run completely.Four, there is good interactivity: program performs different functional modules, such as collecting sample picture module, training classifier module, on-line operation module etc. according to difference input instruction.Operational factor can be changed by user, and diversified Output rusults facilitates Front End to play up display effect, and the mode being stored to database makes this program can be completely independent with Front End, and front end even can adopt different programming languages.
Note, above are only preferred embodiment of the present invention and institute's application technology principle.Skilled person in the art will appreciate that and the invention is not restricted to specific embodiment described here, various obvious change can be carried out for a person skilled in the art, readjust and substitute and can not protection scope of the present invention be departed from.Therefore, although be described in further detail invention has been by above embodiment, the present invention is not limited only to above embodiment, when not departing from the present invention's design, can also comprise other Equivalent embodiments more, and scope of the present invention is determined by appended right.
Claims (5)
1., based on a moving vehicles detection and tracking method for real-time video, it is characterized in that, comprise the steps:
S101, obtain initial sample set:
Image results in file system is classified, is positioned over respectively in different files, obtain initial samples pictures collection; Adopt autoexec to obtain the absolute path list of samples pictures in each file respectively, and give different marks to the file in each file, finally the sample absolute path list of tape label is incorporated in a text;
S102, training preliminary classification device:
The initial sample set training classifier utilizing step S101 to obtain, reads in each sample and mark thereof successively, analyzes the proper vector that its feature obtains each sample, carries out Supervised classification to proper vector, obtain preliminary classification device;
S103, repetitive exercise sorter:
The preliminary classification device utilizing step S102 to obtain is classified to the moving object in video sequence, and the local area image at moving object place is added timestamp name be stored in file system by classification results mark, if user is satisfied to classification results, without the need to carrying out repetitive exercise, if dissatisfied, find out unsatisfied misclassification result, and be added in the sample set file of correct class, method described in step S101 is utilized to obtain new training sample set text, then method described in step S102 is utilized to carry out repetitive exercise, so repeatedly until user is to the satisfied repetitive exercise then completing sorter of classification results, sorter writing in files is saved in file system,
S104, vehicle candidate region generate:
Read in sorter file, detect the moving region image occurred in video sequence successively, the feature analyzing moving region image obtains its proper vector, proper vector is sent into sorter to carry out classifying and obtaining classification results, if result is vehicle, thinks that this moving region image comprises vehicle, complete the generation of vehicle candidate region with this;
S105, vehicle confirm and follow the tracks of:
The vehicle detection result in every two field picture is obtained in step S104, first be benchmark with present frame in real-time video, to all vehicle detection results in each vehicle detection result traversal previous frame of present frame, similarity indices will be met and minimum one of similarity measurement result is considered as the matching result of current goal; Same, take previous frame as benchmark, to the result of each testing result traversal present frame of previous frame, will similarity indices be met and minimum one of similarity measurement result is considered as the matching result of previous frame target; Finally, if previous frame exists and present frame lost target is considered as target disappearance, if the target that present frame occurs does not appear in previous frame, be considered as fresh target and distribute new mark for it, if two targets of present frame and previous frame match each other, be considered as same target and follow the tracks of with same tag.
2. the moving vehicles detection and tracking method based on real-time video according to claim 1, is characterized in that, also comprise:
S106, a wagon flow statistics:
Measurement type is divided into three kinds: the vehicle fleet occurred in picture, sails vehicle fleet into, outgoing vehicles sum; The statistics of vehicle fleet is carried out according to the result of vehicle tracking in step S105, only has and just counts when fresh target occurs; Sail into and add up according to the mode of dummy line with outgoing vehicles, be in line, then think that it is for sailing vehicle into when original state is vehicle's current condition outside line, carrying out counting and its original state being changed is set in line, and wherein, the statistical method of outgoing vehicles is identical with the statistical method of sailing vehicle into.
3. the moving vehicles detection and tracking method based on real-time video according to claim 2, is characterized in that, also comprise:
S107, algorithm are mutual:
Because the local area image of the vehicle target in each two field picture all can be stored in file system, this testing result is comprised the information write server database in camera number, writing time, vehicle target image results path simultaneously; When reaching the vehicle flowrate time interval, by present period data write into Databasce and by counter O reset at every turn; Interactive program can the state in real-time monitoring data storehouse, carries out playing up display result according to data inserting when there being data to insert; Whenever reaching running state of programs daily record interval, by in program current state write journal file, to have good grounds when breaking down, on the other hand, program provides configuration file to revise parameter for user, and user can revise the supplemental characteristic comprising the vehicle flowrate time interval, running state of programs daily record interval, frame picture process dutycycle.
4. according to the moving vehicles detection and tracking method based on real-time video one of claims 1 to 3 Suo Shu, it is characterized in that, described step S105 also comprises: utilize vehicle tracking result to improve vehicle detection effect, the target of tenacious tracking preset times just confirms as vehicle, in this, as vehicle detection result, preserve the image results of vehicle place regional area to file system and by testing result data data inserting storehouse simultaneously.
5. the moving vehicles detection and tracking method based on real-time video according to claim 4, it is characterized in that, described step S105 comprises further: for the result of tenacious tracking, if suddenly disappeared, give predictive compensation and preset frame number, if again occurred with internal object at default frame number, continue to follow the tracks of, otherwise think track rejection.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510210294.5A CN104809437B (en) | 2015-04-28 | 2015-04-28 | A kind of moving vehicles detection and tracking method based on real-time video |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510210294.5A CN104809437B (en) | 2015-04-28 | 2015-04-28 | A kind of moving vehicles detection and tracking method based on real-time video |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104809437A true CN104809437A (en) | 2015-07-29 |
| CN104809437B CN104809437B (en) | 2018-04-13 |
Family
ID=53694248
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510210294.5A Active CN104809437B (en) | 2015-04-28 | 2015-04-28 | A kind of moving vehicles detection and tracking method based on real-time video |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104809437B (en) |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105844222A (en) * | 2016-03-18 | 2016-08-10 | 上海欧菲智能车联科技有限公司 | System and method for front vehicle collision early warning based on visual sense |
| CN106462762A (en) * | 2016-09-16 | 2017-02-22 | 香港应用科技研究院有限公司 | Detection, tracking and positioning of vehicle based on enhanced inverse perspective mapping |
| CN107038423A (en) * | 2017-04-20 | 2017-08-11 | 常州智行科技有限公司 | A kind of vehicle is detected and tracking in real time |
| CN107330365A (en) * | 2017-05-27 | 2017-11-07 | 深圳市美好幸福生活安全***有限公司 | Traffic sign recognition method based on maximum stable extremal region and SVM |
| CN107506753A (en) * | 2017-09-19 | 2017-12-22 | 北京工业大学 | A kind of more wireless vehicle trackings towards dynamic video monitoring |
| CN107909012A (en) * | 2017-10-30 | 2018-04-13 | 北京中科慧眼科技有限公司 | A kind of real-time vehicle tracking detection method and device based on disparity map |
| CN107992900A (en) * | 2017-12-18 | 2018-05-04 | 深圳市盛波光电科技有限公司 | Sample acquiring method, training method, device, medium and the equipment of defects detection |
| CN108830332A (en) * | 2018-06-22 | 2018-11-16 | 安徽江淮汽车集团股份有限公司 | A kind of vision vehicle checking method and system |
| CN108985233A (en) * | 2018-07-19 | 2018-12-11 | 常州智行科技有限公司 | One kind being based on the relevant high-precision wireless vehicle tracking of digital picture |
| CN109191839A (en) * | 2018-09-13 | 2019-01-11 | 江南大学 | Vehicle classification and method of counting based on deep learning |
| CN109726670A (en) * | 2018-12-26 | 2019-05-07 | 浙江捷尚视觉科技股份有限公司 | A method of extracting target detection sample set from video |
| CN110765940A (en) * | 2019-10-22 | 2020-02-07 | 杭州姿感科技有限公司 | Target object statistical method and device |
| CN111046820A (en) * | 2019-12-17 | 2020-04-21 | 上海船舶研究设计院（中国船舶工业集团公司第六0四研究院） | Statistical method and device for vehicles in automobile roll-on-roll-off ship and intelligent terminal |
| CN112130792A (en) * | 2020-09-11 | 2020-12-25 | 安徽中科新辰技术有限公司 | Method for realizing multi-end synchronization of visual views |
| CN112950961A (en) * | 2021-01-27 | 2021-06-11 | 苏州智芯控联信息科技有限公司 | Traffic flow statistical method, device, equipment and storage medium |
| CN112967276A (en) * | 2021-03-31 | 2021-06-15 | 杭州海康威视数字技术股份有限公司 | Object detection method, object detection device, endoscope system, electronic device, and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060109341A1 (en) * | 2002-08-15 | 2006-05-25 | Roke Manor Research Limited | Video motion anomaly detector |
| CN102142090A (en) * | 2011-03-15 | 2011-08-03 | 中国科学技术大学 | Vehicle detection method and system |
| CN103020590A (en) * | 2012-11-20 | 2013-04-03 | 北京航空航天大学深圳研究院 | Vehicle identification system and method based on three-dimensional model and image matching |
| CN103246896A (en) * | 2013-05-24 | 2013-08-14 | 成都方米科技有限公司 | Robust real-time vehicle detection and tracking method |
| CN103258427A (en) * | 2013-04-24 | 2013-08-21 | 北京工业大学 | Urban expressway traffic real-time monitoring system and method based on information physical network |
-
2015
- 2015-04-28 CN CN201510210294.5A patent/CN104809437B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060109341A1 (en) * | 2002-08-15 | 2006-05-25 | Roke Manor Research Limited | Video motion anomaly detector |
| CN102142090A (en) * | 2011-03-15 | 2011-08-03 | 中国科学技术大学 | Vehicle detection method and system |
| CN103020590A (en) * | 2012-11-20 | 2013-04-03 | 北京航空航天大学深圳研究院 | Vehicle identification system and method based on three-dimensional model and image matching |
| CN103258427A (en) * | 2013-04-24 | 2013-08-21 | 北京工业大学 | Urban expressway traffic real-time monitoring system and method based on information physical network |
| CN103246896A (en) * | 2013-05-24 | 2013-08-14 | 成都方米科技有限公司 | Robust real-time vehicle detection and tracking method |
Non-Patent Citations (1)
| Title |
|---|
| 吕万修: "车辆检测跟踪***研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (25)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105844222B (en) * | 2016-03-18 | 2019-07-30 | 上海欧菲智能车联科技有限公司 | The front vehicles collision warning systems and method of view-based access control model |
| CN105844222A (en) * | 2016-03-18 | 2016-08-10 | 上海欧菲智能车联科技有限公司 | System and method for front vehicle collision early warning based on visual sense |
| CN106462762A (en) * | 2016-09-16 | 2017-02-22 | 香港应用科技研究院有限公司 | Detection, tracking and positioning of vehicle based on enhanced inverse perspective mapping |
| CN107038423A (en) * | 2017-04-20 | 2017-08-11 | 常州智行科技有限公司 | A kind of vehicle is detected and tracking in real time |
| CN107038423B (en) * | 2017-04-20 | 2020-03-27 | 常州智行科技有限公司 | Vehicle real-time detection and tracking method |
| CN107330365A (en) * | 2017-05-27 | 2017-11-07 | 深圳市美好幸福生活安全***有限公司 | Traffic sign recognition method based on maximum stable extremal region and SVM |
| CN107506753A (en) * | 2017-09-19 | 2017-12-22 | 北京工业大学 | A kind of more wireless vehicle trackings towards dynamic video monitoring |
| CN107506753B (en) * | 2017-09-19 | 2021-01-05 | 北京工业大学 | Multi-vehicle tracking method for dynamic video monitoring |
| CN107909012A (en) * | 2017-10-30 | 2018-04-13 | 北京中科慧眼科技有限公司 | A kind of real-time vehicle tracking detection method and device based on disparity map |
| CN107909012B (en) * | 2017-10-30 | 2022-03-18 | 北京中科慧眼科技有限公司 | Real-time vehicle tracking detection method and device based on disparity map |
| CN107992900A (en) * | 2017-12-18 | 2018-05-04 | 深圳市盛波光电科技有限公司 | Sample acquiring method, training method, device, medium and the equipment of defects detection |
| CN107992900B (en) * | 2017-12-18 | 2020-09-11 | 深圳市盛波光电科技有限公司 | Sample acquisition method, training method, device, medium and equipment for defect detection |
| CN108830332A (en) * | 2018-06-22 | 2018-11-16 | 安徽江淮汽车集团股份有限公司 | A kind of vision vehicle checking method and system |
| CN108985233A (en) * | 2018-07-19 | 2018-12-11 | 常州智行科技有限公司 | One kind being based on the relevant high-precision wireless vehicle tracking of digital picture |
| CN108985233B (en) * | 2018-07-19 | 2021-08-03 | 常州智行科技有限公司 | High-precision vehicle tracking method based on digital image correlation |
| CN109191839A (en) * | 2018-09-13 | 2019-01-11 | 江南大学 | Vehicle classification and method of counting based on deep learning |
| CN109726670A (en) * | 2018-12-26 | 2019-05-07 | 浙江捷尚视觉科技股份有限公司 | A method of extracting target detection sample set from video |
| CN110765940A (en) * | 2019-10-22 | 2020-02-07 | 杭州姿感科技有限公司 | Target object statistical method and device |
| CN111046820A (en) * | 2019-12-17 | 2020-04-21 | 上海船舶研究设计院（中国船舶工业集团公司第六0四研究院） | Statistical method and device for vehicles in automobile roll-on-roll-off ship and intelligent terminal |
| CN112130792A (en) * | 2020-09-11 | 2020-12-25 | 安徽中科新辰技术有限公司 | Method for realizing multi-end synchronization of visual views |
| CN112130792B (en) * | 2020-09-11 | 2023-12-15 | 安徽中科新辰技术有限公司 | Method for realizing multi-terminal synchronization of visual view |
| CN112950961A (en) * | 2021-01-27 | 2021-06-11 | 苏州智芯控联信息科技有限公司 | Traffic flow statistical method, device, equipment and storage medium |
| CN112950961B (en) * | 2021-01-27 | 2022-07-08 | 苏州智芯控联信息科技有限公司 | Traffic flow statistical method, device, equipment and storage medium |
| CN112967276A (en) * | 2021-03-31 | 2021-06-15 | 杭州海康威视数字技术股份有限公司 | Object detection method, object detection device, endoscope system, electronic device, and storage medium |
| CN112967276B (en) * | 2021-03-31 | 2023-09-05 | 杭州海康威视数字技术股份有限公司 | Object detection method, object detection device, endoscope system, electronic device, and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104809437B (en) | 2018-04-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104809437A (en) | Real-time video based vehicle detecting and tracking method | |
| Unzueta et al. | Adaptive multicue background subtraction for robust vehicle counting and classification | |
| Messelodi et al. | A computer vision system for the detection and classification of vehicles at urban road intersections | |
| CN103336954B (en) | A kind of TV station symbol recognition method and apparatus in video | |
| Keller et al. | A new benchmark for stereo-based pedestrian detection | |
| CN105513349B (en) | Mountainous area highway vehicular events detection method based on double-visual angle study | |
| Lin et al. | A Real‐Time Vehicle Counting, Speed Estimation, and Classification System Based on Virtual Detection Zone and YOLO | |
| Mallikarjuna et al. | Traffic data collection under mixed traffic conditions using video image processing | |
| Bach et al. | Deep convolutional traffic light recognition for automated driving | |
| CN104134222A (en) | Traffic flow monitoring image detecting and tracking system and method based on multi-feature fusion | |
| CN107886055A (en) | A kind of retrograde detection method judged for direction of vehicle movement | |
| JP2011034244A (en) | Image analysis apparatus, image analysis method and program | |
| Han et al. | Aerial image change detection using dual regions of interest networks | |
| CN104615986A (en) | Method for utilizing multiple detectors to conduct pedestrian detection on video images of scene change | |
| CN105260749A (en) | Real-time target detection method based on oriented gradient two-value mode and soft cascade SVM | |
| CN102254428A (en) | Traffic jam detection method based on video processing | |
| CN111079621B (en) | Method, device, electronic equipment and storage medium for detecting object | |
| JP2011059898A (en) | Image analysis apparatus and method, and program | |
| CN109740609A (en) | A kind of gauge detection method and device | |
| CN108288020A (en) | Video shelter detecting system based on contextual information and method | |
| Yaghoobi Ershadi et al. | Vehicle tracking and counting system in dusty weather with vibrating camera conditions | |
| CN104123714A (en) | Optimal target detection scale generation method in people flow statistics | |
| Diego et al. | Vision-based road detection via on-line video registration | |
| EP3761228A1 (en) | Computer-implemented method | |
| Gao et al. | Anomaly detection of trackside equipment based on GPS and image matching |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |