CN103400113A - Method for detecting pedestrian on expressway or in tunnel based on image processing - Google Patents
Method for detecting pedestrian on expressway or in tunnel based on image processing Download PDFInfo
- Publication number
- CN103400113A CN103400113A CN2013102907396A CN201310290739A CN103400113A CN 103400113 A CN103400113 A CN 103400113A CN 2013102907396 A CN2013102907396 A CN 2013102907396A CN 201310290739 A CN201310290739 A CN 201310290739A CN 103400113 A CN103400113 A CN 103400113A
- Authority
- CN
- China
- Prior art keywords
- image
- area
- pedestrian
- foreground target
- width
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Landscapes
- Image Analysis (AREA)
- Image Processing (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention relates to the technical field of traffic detection and discloses a method for detecting a pedestrian on an expressway or in a tunnel based on image processing. The method comprises the following steps: 1) acquiring an image from a video obtained by a camera on the expressway or in the tunnel according to a preset frame rate; 2) extracting a foreground target from the picture; 3) obtaining the contour area of the foreground target and the width-to-height ratio of the contour circumscribed rectangle of the foreground target; 4) primarily judging whether a pedestrian target exists in the current frame of image or not according to the contour area of the foreground target and the width-to-height ratio of the contour circumscribed rectangle of the foreground target; 5) repeating steps 1-4) to obtain the primary judgment results of multiple frames of images and finally judging whether the pedestrian target exists or not. After one frame of image is primarily judged according to the contour area and the width-to-height ratio of the circumscribed rectangle, whether the pedestrian target exists or not is finally judged in consideration of the detection results of multiple frames of images, so that the pedestrian target can be accurately and effectively detected. The method is low in operation expense and has strong real-time performance.
Description
Technical field
The present invention relates to traffic detection technique field, relate in particular to a kind of freeway tunnel pedestrian detection method of processing based on image.
Background technology
Along with the fast development of China's expressway construction, freeway tunnel puts into effect in a large number.When freeway tunnel made the efficient UNICOM of traffic, due to its special environment, operation security was subject to common concern.Wherein, pedestrians disobeying traffic rule is the important risk factors of freeway tunnel traffic hazard, is also the important feature of China's traffic accidents.Pedestrians disobeying traffic rule in freeway tunnel very easily causes vehicle and pedestrian impact, brings very big hidden danger for pedestrian's life security, even can cause more serious traffic hazard.Therefore pedestrians disobeying traffic rule is freeway tunnel traffic safety problem demanding prompt solution.
At present, utilize the video monitoring system of freeway tunnel, realized the Real Time Monitoring to pedestrian's situation.But the discovery to the pedestrian, remain by the staff and observe sequence of video images, realizes unrealized pedestrian's automatic detection by manual supervisory mode.Therefore, how to utilize video detection technology automatically to detect in real time the pedestrian, for the traffic operation and management person, make in real time management decision, the traffic circulation security level in Improving Expressway tunnel has great significance.
In prior art, the pedestrian detection method of processing based on image mainly contains method based on manikin, based on the method for template matches with based on the method for statistical classification.Based on the solution procedure more complicated of model in the method for manikin, the expense degree is large.Be difficult to construct enough templates based on the method for template matches due to pedestrian's polymorphism, also be difficult to realize.Based on the method for statistical classification, need to extract a plurality of features and training great amount of samples data, realize loaded down with trivial detailsly, complexity is large." based on the pedestrian detection method of video monitoring " (publication number: CN101887524) of Hunan Chuanghe Manufacturing Co., Ltd's application, utilize expansion histogram of gradients feature and Adaboost algorithm to carry out the fast detecting pedestrian, then utilize histogram of gradients feature and support vector machine further to identify the pedestrian that checking detects previously.This method need to be extracted the histogram of gradients feature of image, video image quality is very not clear, lighting is brighter and target hour, the pedestrian target histogram of gradients feature of obtaining is very undesirable, can't realize the pedestrian target that occurs in tunnel is detected." a kind of pedestrian detection method based on small echo somatotype feature " (publication number: CN101630369) of Nanjing Aero-Space University's application, by sample set is carried out repeatedly two-dimensional wavelet transformation, be used for training soft support vector machine thereby extract wavelet fractal characteristic, obtain discriminant function and realize pedestrian detection.This method need to be carried out repeatedly wavelet transformation to video image, calculates loaded down with trivial detailsly, and expense is very large, and, when the target range camera position is far away and target when very little, can't realize the detection to pedestrian target after wavelet transformation.(publication number: CN102147869) utilize the prospect that contour feature and the analysis of pedestrian level prior model obtain and obtain preliminary pedestrian detection result, recycling pedestrian pattern recognition classifier device obtains final pedestrian detection result to " based on the pedestrian detection method of prospect somatotype and pattern-recognition " of Shanghai Communications University application.This method can obtain the Preliminary detection result rapidly, but in the freeway tunnel of reality, houselights and lights of vehicle are very large to the interference of testing result, by the pattern recognition classifier device, the testing result of single-frame images are classified, and easily produce a large amount of flase drops.
Just because of this, need a kind of can be accurately, in real time, effective pedestrian detection method, provide the data message of use for the traffic safety of freeway tunnel.
Summary of the invention
In view of this, the invention provides a kind of freeway tunnel pedestrian detection method of processing based on image, the computing expense is little, real-time, and accuracy is high.
The present invention solves the problems of the technologies described above by following technological means:
Freeway tunnel pedestrian detection method based on image is processed, comprise the steps:
1) obtain image by default frame per second from the video that the freeway tunnel camera obtains;
2) extract foreground target from picture;
3) obtain foreground target contour area and foreground target profile boundary rectangle the ratio of width to height;
4) tentatively judge in current frame image whether have pedestrian target according to foreground target contour area and foreground target profile boundary rectangle the ratio of width to height;
5) repeating step 1-4), obtain the preliminary judged result of multiple image, judge whether to exist pedestrian target, specifically comprise the steps:
51) calculate the accumulation judged result C of k frame
k Wherein Num represents the frame number of statistics, wherein k continuously〉Num, Judge
k, for the preliminary judged result in step 4), tentatively be judged as while having pedestrian target Judge
kBe 1, otherwise be 0;
53) give and work as V
kDuring 〉=Alert, finally judge existing people; Wherein Alert ∈ [0,1], be the pedestrian detection alarm threshold value.
Further, in described step 1), in the image acceptance of the bid regular inspection that obtains, survey zone; Step 2-5) only surveyed area is processed.
Further, described step 2), specifically comprise the steps:
21) adopt double method of difference to process the image that obtains: it is poor that the k+1 frame of video image and k frame are done, and it is poor that k+2 frame and k+1 frame are done;
22) image after difference processing is carried out binaryzation;
23) two width images after binaryzation are carried out the logical OR computing;
24) image after logical operation is carried out morphology and process, obtain foreground target.
Further, described step 22) in, choosing of binary-state threshold adopts maximum variance between clusters to determine.
Further, step 24) in, adopt the opening operation method to carry out morphology and process.
Further, described step 3), specifically comprise the steps:
31) obtain the contour area Area of foreground target;
32) obtain the boundary rectangle of foreground target profile, calculate the ratio of width to height W_H of described boundary rectangle.Adopt formula
Wherein Wide represents the width of foreground target boundary rectangle, and Height represents the height of foreground target boundary rectangle.
Further, described step 4) specifically comprises the steps:
41) judge whether the foreground target contour area meets pedestrian's feature: judge whether contour area Area meets pre-conditioned: Area_min≤Area≤Area_max, wherein Area_min represents the lower limit of default contour area, Area_max represents the upper limit of default contour area, and 0<Area_min<Area_max;
Whether the ratio of width to height that 42) judges foreground target profile boundary rectangle meets pedestrian's feature: whether the ratio of width to height W_H that judges boundary rectangle meets: W_H_min≤W_H≤W_H_max, wherein W_H_min represents the default lower limit of boundary rectangle the ratio of width to height, W_H_max represents the preset upper limit of boundary rectangle the ratio of width to height, and 0<W_H_min<W_H_max;
43) draw the preliminary judged result Judge of k frame pedestrian target
k: foreground target contour area and boundary rectangle the ratio of width to height all meet 41), 42) judgement the time, Judge
kBe 1, otherwise be 0.
The freeway tunnel pedestrian detection method of processing based on image of the present invention, when the judgement pedestrian target, only need surveyed area is demarcated and related to two characteristic parameters of contour area boundary rectangle the ratio of width to height, and objective contour area and boundary rectangle the ratio of width to height all can obtain exactly from video image, therefore the present invention is when adopting criterion based on contour area boundary rectangle the ratio of width to height to carry out pedestrian detection, can detect accurately and efficiently pedestrian target, and the computing expense is little, and is real-time.
Embodiment
Fig. 1 shows the schematic flow sheet of the freeway tunnel pedestrian detection method of processing based on image.
Embodiment
Below with reference to accompanying drawing, the present invention is described in detail.
Referring to Fig. 1, the freeway tunnel pedestrian detection method based on image is processed, comprise the steps:
1) obtain image by default frame per second from the video that the freeway tunnel camera obtains, and in the image acceptance of the bid regular inspection that obtains, survey zone; Step 2-5) only surveyed area is processed; Demarcate surveyed area the image range of processing is diminished, thereby reduce the data volume of processing, improve the high efficiency of algorithm.Demarcate in addition surveyed area and can also avoid the impact of some disturbing factors on detecting;
2) extract foreground target from picture; Specifically comprise the steps:
21) adopt double method of difference to process the image that obtains: it is poor that the k+1 frame of video image and k frame are done, and it is poor that k+2 frame and k+1 frame are done; In prior art, main target extraction method is background subtraction point-score and frame differential method, impact due to lights of vehicle in tunnel and houselights, the effect of background modeling can be interfered, therefore the background subtraction point-score can't obtain good foreground target, frame differential method can extract good moving target and can avoid the impact of change of background, but common frame differential method can make the moving target of extraction imperfect, so the present embodiment adopts double method of difference, can obtain comparatively complete foreground target like this;
22) image after difference processing is carried out binaryzation, choosing of binary-state threshold adopts maximum variance between clusters to determine, adopts maximum variance between clusters to choose image segmentation threshold and can obtain optimum efficiency;
23) two width images after binaryzation are carried out the logical OR computing;
24) image of the method that adopts opening operation after with logical operation carries out morphology to be processed, and removes less noise and can fill some spaces, the acquisition foreground target.
3) obtain foreground target contour area and foreground target profile boundary rectangle the ratio of width to height, specifically comprise the steps:
31) obtain the contour area Area of foreground target;
32) obtain the boundary rectangle of foreground target profile, calculate the ratio of width to height W_H of described boundary rectangle.Adopt formula
Wherein Wide represents the width of foreground target boundary rectangle, and Height represents the height of foreground target boundary rectangle.
Due in tunnel lamplit strong and weak different, vary in color, the color histogram feature that pedestrian detection is commonly used is very unstable in the different scenes of freeway tunnel, and the contour area of target and boundary rectangle the ratio of width to height are two more stable features, can characterize well pedestrian target.
4) tentatively judge in current frame image whether have pedestrian target, specifically comprise the steps: according to foreground target contour area and foreground target profile boundary rectangle the ratio of width to height
41) judge whether the foreground target contour area meets pedestrian's feature: judge whether contour area Area meets pre-conditioned: Area_min≤Area≤Area_max, wherein Area_min represents the lower limit of default contour area, Area_max represents the upper limit of default contour area, and 0<Area_min<Area_max;
Whether the ratio of width to height that 42) judges foreground target profile boundary rectangle meets pedestrian's feature: whether the ratio of width to height W_H that judges boundary rectangle meets: W_H_min≤W_H≤W_H_max, wherein W_H_min represents the default lower limit of boundary rectangle the ratio of width to height, W_H_max represents the preset upper limit of boundary rectangle the ratio of width to height, and 0<W_H_min<W_H_max;
43) draw the preliminary judged result Judge of k frame pedestrian target
k: foreground target contour area and boundary rectangle the ratio of width to height all meet 41), 42) judgement the time, Judge
kBe 1, otherwise be 0.
5) repeating step 1-4), obtain the preliminary judged result of multiple image, judge whether to exist pedestrian target, specifically comprise the steps:
51) calculate the accumulation judged result C of k frame
k Wherein Num represents the frame number of statistics, wherein k continuously〉Num, Judge
k, for the preliminary judged result in step 4), tentatively be judged as while having pedestrian target Judge
kBe 1, otherwise be 0;
53) give and work as V
kDuring 〉=Alert, finally judge existing people; Wherein Alert ∈ [0,1], be the pedestrian detection alarm threshold value.
When final judgement have the pedestrian by the time can carry out alarm to the relevant staff, the generation that the prompting relevant staff takes necessary measure to try to forestall traffic accidents.
Video image is that three-dimensional world is transformed in two-dimensional coordinate, the vehicle target that travels in freeway tunnel in road can project to the pedestrian detection zone, lights of vehicle also can disturb the pedestrian detection zone, so by single-frame images, detect the pedestrian, can cause a lot of flase drops.Need the regular hour due to the pedestrian during by video monitoring regional, and vehicle is usually all very short by the time that needs, so can, in conjunction with the preliminary judged result of continuous multiple frames, finally judge whether to occur the pedestrian.
Explanation is finally, above embodiment is only unrestricted in order to technical scheme of the present invention to be described, although with reference to preferred embodiment, the present invention is had been described in detail, those of ordinary skill in the art is to be understood that, can modify or be equal to replacement technical scheme of the present invention, and not breaking away from aim and the scope of technical solution of the present invention, it all should be encompassed in the middle of claim scope of the present invention.
Claims (7)
1. the freeway tunnel pedestrian detection method of processing based on image, is characterized in that: comprise the steps:
1) obtain image by default frame per second from the video that the freeway tunnel camera obtains;
2) extract foreground target from picture;
3) obtain foreground target contour area and foreground target profile boundary rectangle the ratio of width to height;
4) tentatively judge in current frame image whether have pedestrian target according to foreground target contour area and foreground target profile boundary rectangle the ratio of width to height;
5) repeating step 1-4), obtain the preliminary judged result of multiple image, judge whether to exist pedestrian target, specifically comprise the steps:
51) calculate the accumulation judged result C of k frame
k Wherein Num represents the frame number of statistics, wherein k continuously〉Num, Judge
k, for the preliminary judged result in step 4), tentatively be judged as while having pedestrian target Judge
kBe 1, otherwise be 0.
53) give and work as V
kDuring 〉=Alert, finally judge existing people; Wherein Alert ∈ [0,1], be the pedestrian detection alarm threshold value.
2. the freeway tunnel pedestrian detection method of processing based on image as claimed in claim 1, is characterized in that: in described step 1), in the image acceptance of the bid regular inspection that obtains, survey zone; Step 2-5) only surveyed area is processed.
3. the freeway tunnel pedestrian detection method of processing based on image as claimed in claim 1, is characterized in that: described step 2), specifically comprise the steps:
21) adopt double method of difference to process the image that obtains: it is poor that the k+1 frame of video image and k frame are done, and it is poor that k+2 frame and k+1 frame are done;
22) image after difference processing is carried out binaryzation;
23) two width images after binaryzation are carried out the logical OR computing;
24) image after logical operation is carried out morphology and process, obtain foreground target.
4. the freeway tunnel pedestrian detection method of processing based on image as claimed in claim 3, it is characterized in that: described step 22), choosing of binary-state threshold adopts maximum variance between clusters to determine.
5. the freeway tunnel pedestrian detection method of processing based on image as claimed in claim 3 is characterized in that:
Step 24) in, adopt the opening operation method to carry out morphology and process.
6. the freeway tunnel pedestrian detection method of processing based on image as described in any one in claim 1-5, it is characterized in that: described step 3) specifically comprises the steps:
31) obtain the contour area Area of foreground target;
32) obtain the boundary rectangle of foreground target profile, calculate the ratio of width to height W_H of described boundary rectangle.Adopt formula
Wherein Wide represents the width of foreground target boundary rectangle, and Height represents the height of foreground target boundary rectangle.
7. the freeway tunnel pedestrian detection method of processing based on image as claimed in claim 6, it is characterized in that: described step 4) specifically comprises the steps:
41) judge whether the foreground target contour area meets pedestrian's feature: judge whether contour area Area meets pre-conditioned: Area_min≤Area≤Area_max, wherein Area_min represents the lower limit of default contour area, Area_max represents the upper limit of default contour area, and 0<Area_min<Area_max;
Whether the ratio of width to height that 42) judges foreground target profile boundary rectangle meets pedestrian's feature: whether the ratio of width to height W_H that judges boundary rectangle meets: W_H_min≤W_H≤W_H_max, wherein W_H_min represents the default lower limit of boundary rectangle the ratio of width to height, W_H_max represents the preset upper limit of boundary rectangle the ratio of width to height, and 0<W_H_min<W_H_max;
43) draw the preliminary judged result Judge of k frame pedestrian target
k: foreground target contour area and boundary rectangle the ratio of width to height all meet 41), 42) judgement the time, Judge
kBe 1, otherwise be 0.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310290739.6A CN103400113B (en) | 2013-07-10 | 2013-07-10 | Freeway tunnel pedestrian detection method based on image procossing |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310290739.6A CN103400113B (en) | 2013-07-10 | 2013-07-10 | Freeway tunnel pedestrian detection method based on image procossing |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103400113A true CN103400113A (en) | 2013-11-20 |
CN103400113B CN103400113B (en) | 2016-08-24 |
Family
ID=49563731
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310290739.6A Active CN103400113B (en) | 2013-07-10 | 2013-07-10 | Freeway tunnel pedestrian detection method based on image procossing |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103400113B (en) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104239865A (en) * | 2014-09-16 | 2014-12-24 | 宁波熵联信息技术有限公司 | Pedestrian detecting and tracking method based on multi-stage detection |
CN105338304A (en) * | 2014-08-13 | 2016-02-17 | 南宁市锋威科技有限公司 | Tunnel operation safety incident detection system based on video recognition |
CN105866790A (en) * | 2016-04-07 | 2016-08-17 | 重庆大学 | Laser radar barrier identification method and system taking laser emission intensity into consideration |
CN106682566A (en) * | 2015-11-09 | 2017-05-17 | 富士通株式会社 | Traffic accident detection method, traffic accident detection device and electronic device |
CN107818651A (en) * | 2017-10-27 | 2018-03-20 | 华润电力技术研究院有限公司 | A kind of illegal cross-border warning method and device based on video monitoring |
CN108563977A (en) * | 2017-12-18 | 2018-09-21 | 华南理工大学 | A kind of the pedestrian's method for early warning and system of expressway entrance and exit |
CN110276742A (en) * | 2019-05-07 | 2019-09-24 | 平安科技(深圳)有限公司 | Tail light for train monitoring method, device, terminal and storage medium |
CN110852253A (en) * | 2019-11-08 | 2020-02-28 | 杭州宇泛智能科技有限公司 | Ladder control scene detection method and device and electronic equipment |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20110255741A1 (en) * | 2010-02-05 | 2011-10-20 | Sang-Hack Jung | Method and apparatus for real-time pedestrian detection for urban driving |
CN102542289A (en) * | 2011-12-16 | 2012-07-04 | 重庆邮电大学 | Pedestrian volume statistical method based on plurality of Gaussian counting models |
CN102768726A (en) * | 2011-05-06 | 2012-11-07 | 香港生产力促进局 | Pedestrian detection method for preventing pedestrian collision |
CN103049751A (en) * | 2013-01-24 | 2013-04-17 | 苏州大学 | Improved weighting region matching high-altitude video pedestrian recognizing method |
-
2013
- 2013-07-10 CN CN201310290739.6A patent/CN103400113B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20110255741A1 (en) * | 2010-02-05 | 2011-10-20 | Sang-Hack Jung | Method and apparatus for real-time pedestrian detection for urban driving |
CN102768726A (en) * | 2011-05-06 | 2012-11-07 | 香港生产力促进局 | Pedestrian detection method for preventing pedestrian collision |
CN102542289A (en) * | 2011-12-16 | 2012-07-04 | 重庆邮电大学 | Pedestrian volume statistical method based on plurality of Gaussian counting models |
CN103049751A (en) * | 2013-01-24 | 2013-04-17 | 苏州大学 | Improved weighting region matching high-altitude video pedestrian recognizing method |
Non-Patent Citations (4)
Title |
---|
JUNFENG GE ET AL: "real-time pedestrian detection and tracking at nighttime for driver-assistance systems", 《IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS》 * |
庄家俊,刘琼: "面向辅助驾驶的夜间行人检测方法", 《华南理工大学学报(自然科学版)》 * |
张磊: "动态目标图像识别与跟踪技术研究", 《中国优秀硕士学位论文全文数据库信息科技辑2012年》 * |
梁成: "基于视频的行人检测技术研究", 《中国优秀硕士学位论文全文数据库信息科技辑2012年》 * |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105338304A (en) * | 2014-08-13 | 2016-02-17 | 南宁市锋威科技有限公司 | Tunnel operation safety incident detection system based on video recognition |
CN104239865A (en) * | 2014-09-16 | 2014-12-24 | 宁波熵联信息技术有限公司 | Pedestrian detecting and tracking method based on multi-stage detection |
CN104239865B (en) * | 2014-09-16 | 2017-04-12 | 宁波熵联信息技术有限公司 | Pedestrian detecting and tracking method based on multi-stage detection |
CN106682566A (en) * | 2015-11-09 | 2017-05-17 | 富士通株式会社 | Traffic accident detection method, traffic accident detection device and electronic device |
CN105866790A (en) * | 2016-04-07 | 2016-08-17 | 重庆大学 | Laser radar barrier identification method and system taking laser emission intensity into consideration |
CN105866790B (en) * | 2016-04-07 | 2018-08-10 | 重庆大学 | A kind of laser radar obstacle recognition method and system considering lasing intensity |
CN107818651A (en) * | 2017-10-27 | 2018-03-20 | 华润电力技术研究院有限公司 | A kind of illegal cross-border warning method and device based on video monitoring |
CN108563977A (en) * | 2017-12-18 | 2018-09-21 | 华南理工大学 | A kind of the pedestrian's method for early warning and system of expressway entrance and exit |
CN110276742A (en) * | 2019-05-07 | 2019-09-24 | 平安科技(深圳)有限公司 | Tail light for train monitoring method, device, terminal and storage medium |
CN110276742B (en) * | 2019-05-07 | 2023-10-10 | 平安科技(深圳)有限公司 | Train tail lamp monitoring method, device, terminal and storage medium |
CN110852253A (en) * | 2019-11-08 | 2020-02-28 | 杭州宇泛智能科技有限公司 | Ladder control scene detection method and device and electronic equipment |
Also Published As
Publication number | Publication date |
---|---|
CN103400113B (en) | 2016-08-24 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103400113A (en) | Method for detecting pedestrian on expressway or in tunnel based on image processing | |
CN103116985B (en) | Detection method and device of parking against rules | |
WO2019196131A1 (en) | Method and apparatus for filtering regions of interest for vehicle-mounted thermal imaging pedestrian detection | |
CN102298781B (en) | Motion shadow detection method based on color and gradient characteristics | |
CN110517288A (en) | Real-time target detecting and tracking method based on panorama multichannel 4k video image | |
CN104036262B (en) | A kind of method and system of LPR car plates screening identification | |
CN105469105A (en) | Cigarette smoke detection method based on video monitoring | |
CN103617410A (en) | Highway tunnel parking detection method based on video detection technology | |
CN103077384A (en) | Method and system for positioning and recognizing vehicle logo | |
CN103208185A (en) | Method and system for nighttime vehicle detection on basis of vehicle light identification | |
Zhang et al. | A multi-feature fusion based traffic light recognition algorithm for intelligent vehicles | |
CN102542289A (en) | Pedestrian volume statistical method based on plurality of Gaussian counting models | |
Xu et al. | Real-time pedestrian detection based on edge factor and Histogram of Oriented Gradient | |
CN108280409B (en) | Large-space video smoke detection method based on multi-feature fusion | |
CN111553214B (en) | Method and system for detecting smoking behavior of driver | |
CN103077423A (en) | Crowd quantity estimating, local crowd clustering state and crowd running state detection method based on video stream | |
CN103279737A (en) | Fight behavior detection method based on spatio-temporal interest point | |
CN103729858A (en) | Method for detecting article left over in video monitoring system | |
CN103034852A (en) | Specific color pedestrian detecting method in static video camera scene | |
CN105354857B (en) | A kind of track of vehicle matching process for thering is viaduct to block | |
CN103679146A (en) | Safety belt intelligent detection method based on high-pass filter and Hough conversion | |
CN104616006A (en) | Surveillance video oriented bearded face detection method | |
CN105554462A (en) | Remnant detection method | |
WO2014079058A1 (en) | Method and system for processing video image | |
CN103049788A (en) | Computer-vision-based system and method for detecting number of pedestrians waiting to cross crosswalk |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
TR01 | Transfer of patent right | ||
TR01 | Transfer of patent right |
Effective date of registration: 20201216 Address after: 402460 station No.14, no.6, 10th floor, innovation and development center, No.19 Lingfang Avenue, Changzhou street, Rongchang District, Chongqing Patentee after: Chongqing kezhiyuan Technology Co.,Ltd. Address before: 400030 No. 174 Sha Jie street, Shapingba District, Chongqing Patentee before: Chongqing University |