CN103268706A - Method for detecting vehicle queue length based on local variance - Google Patents
Method for detecting vehicle queue length based on local variance Download PDFInfo
- Publication number
- CN103268706A CN103268706A CN2013101347077A CN201310134707A CN103268706A CN 103268706 A CN103268706 A CN 103268706A CN 2013101347077 A CN2013101347077 A CN 2013101347077A CN 201310134707 A CN201310134707 A CN 201310134707A CN 103268706 A CN103268706 A CN 103268706A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- result
- vehicle queue
- queue
- roi
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Landscapes
- Image Processing (AREA)
- Image Analysis (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention disclose a method for detecting a vehicle queue length based on a local variance, which is applied to the field of intelligent transportation, and particularly relates to a digital image processing and computer visual technique. The method comprises the following steps: putting forward an improved local variance method for performing vehicle detection; putting forward a combination of the improved local variance method and an LBP (Length between Perpendiculars) algorithm to remove vehicle shadows; and putting forward a method for reducing vehicle queue length calculation errors. The method for detecting the vehicle queue length has a small calculation quantity, and the problem of error detection of vehicles caused by vehicle shadows of lanes during multi-lane detection is solved. Due to the adoption of the method, functions of real-time control of traffic lights, measurement of traffic flow control indexes and the like can be realized; and the method plays important roles in maintaining traffic safety and urban security, increasing the road utilization ratios, reducing traffic jams, realizing automatic traffic management and intelligent city, and the like.
Description
Technical field
The inventive method is applied to intelligent transportation field, is specifically related to Digital Image Processing, computer vision technique.
Background technology
Along with the continuous appearance of high-performance computer and video appliances and human to the automatic analysis and understanding of video information require growing, the analysis of video information develops into an important and popular research topic in the computer vision research field in recent years gradually, and has attracted more and more researchers.
Along with the progressively quickening of Urbanization in China in recent years, it is more and more serious that traffic congestion has become.Traffic congestion makes the traffic hazard incidence uprise, and has not only aggravated environmental pollution, causes the wasting of resources, and the trip of returning people brings great inconvenience.In the face of the traffic of more and more blocking up and limited soil economic resources, intelligent transportation system ITS (Intelligent Traffic System) arises at the historic moment.And the vehicle queue length of right-angled intersection is one of important composition among the ITS, can estimate the current intelligence of whole road by the length of vehicle queue behind the stop line, and can control in real time traffic light cycles by this parameter.
The vehicle queue detection method is based on technology such as Digital Image Processing, computer visions, and vehicle image or the video sequence of shot by camera are analyzed, and finishes the process that vehicle queue detects thereby obtain successive vehicles length.Can realize the real-time control of traffic lights, functions such as magnitude of traffic flow control index measurement by some method for subsequent processing.For safeguarding traffic safety and urban public security, improve road utilization rate, reduce traffic congestion, realize that there is realistic meanings in traffic automation management and wisdom city etc.But existing successive vehicles length detecting method ubiquity vehicle detecting algorithm complexity, calculated amount is big, and the vehicle shadow of not considering adjacent lane when multilane the detects problem that causes flase drop to survey to vehicle detection.
Summary of the invention
The object of the invention is to overcome the deficiencies in the prior art, and a kind of successive vehicles length detecting method based on local variance is provided.
The technical solution of the present invention method is characterized by:
A kind of method that detects the vehicle queue length of road junction is characterized in that, comprises the steps:
(1) adopt monocular-camera to obtain traffic video in real time at road junction.
(2) video that video camera is imported in real time carries out the virtual coil demarcation, handles knowledge according to image and judges lights state.
Image when gathering a frame vehicle queue when (3) red light becomes green light.
(4) region of interest ROI (Region Of Interest) being carried out vehicle queue detects, namely utilize improved local variance and improved LBP algorithm to combine and remove the shade of vehicle queue, improve the precision that vehicle queue detects, obtain vehicle queue prospect binary map in the ROI.
(5) camera calibration is carried out in the ROI zone, the computing scale chi.By homography matrix H, obtain the physical length that represents under the alive boundary of the center line epigraph two-dimensional image vegetarian refreshments coordinate system in the ROI, with each pixel and actual range one by one the table of corresponding acquisition be rule.
(6) integrating step (4) and both results of step (5), the tail of the queue that arrives by the vehicle foreground detection and team's head portrait vegetarian refreshments, and obtain the physical length L1 of road junction vehicle queue according to the actual range of rule correspondence.For setting height(from bottom) H and the overall height h that overcomes video camera can produce picture vision blind area, improve the accuracy rate that vehicle queue length calculates, the method that reply solves is:
Calculate vehicle queue length L=L1-l according to formula, wherein, L is vehicle commander's degree in the reality, d be video camera to the distance of stop line, the height of car is h, and triangular relationship h/H=l/ (d+L1) is arranged, calculate l=h (d+L1)/H, then final successive vehicles length L=L1 (1-h/H)-hd/H.
In (4) step, the method for vehicle queue detection specific implementation is in the described ROI:
Step (1): lane line is demarcated, delimited out zone, track ROI, get result and be " 1 ", for step (21), the follow-up processing respectively of step (22);
Step (21): in the hsv color space, adopt tone component H to operate; Result " 1 " is calculated each pixel in every 5*5 matrix-block tone component H square value average E{I (i, j)
2And the square value E of mean value (i, j)
2, again according to formula: D (i, j)=E{I (i, j)
2}-E (i, j)
2Calculate the variance D (i of the tone component H of each pixel, j), picture codomain scope after handling like this is bigger, variance D (i with all pixel tone component H, j) normalize in [0-1] scope, and adopting dual threshold method (this algorithm adopts T1=0.88 and T2=0.95) that picture is carried out binaryzation respectively, result is respectively " 3 " and " 2 ";
Again result " 2 " is corroded operation and obtain result " 4 ";
Simultaneously, step (22): select the tone component H of result " 1 " in the step (1) to handle, adopt improved LBP algorithm to extract vehicle formation texture information and obtain result " 5 ",
Described improved LBP algorithm calculates according to following formula:
Step (3): step (2-1) result " 5 " and step (22) result " 4 " are carried out and operation, remove vehicle shadow, obtain result " 6 ";
Step (4): step (2-1) is handled back result " 3 " and step (3) handle afterwards that result " 6 " carries out or operates, adopt the morphology expansive working again, obtain vehicle formation prospect, and to prospect projection on the ROI center line, and definite head of the queue and tail of the queue place pixel.
Through behind the aforesaid operations, just can extract the vehicle queue prospect, and remove the influence of vehicle shadow.
The present invention to state of the art innovative point:
1) proposes improved local variance method and carry out vehicle detection.
2) improved local variance and the improved LBP algorithm removal vehicle shadow that combines is proposed.
3) proposition reduces the method for the car queue length error of calculation.
Successive vehicles length detecting method calculated amount of the present invention is little, and the vehicle shadow that has solved the track when multilane detects causes the problem of flase drop survey to vehicle detection.This application implementation, can realize traffic lights real-time control, measure function such as magnitude of traffic flow control index, for safeguarding traffic safety and urban public security, improve road utilization rate, reduce traffic congestion, realize that traffic automation management and wisdom city etc. are significant.
Description of drawings
The framework that Fig. 1 vehicle queue length detects.
Fig. 2 is the image acquisition process flow diagram.
Fig. 3 is vehicle length detection process flow diagram.
Fig. 4 is the vehicle queue length calculation flow chart.
Fig. 5 is the error note that measuring vehicle formation method produces.
Embodiment
As shown in Figure 1: carry out the process that whole vehicle queue length detects.
Traffic control system is to be based upon on the basis that vehicle queue detects to obtaining with presenting of final vehicle queue length.Detailed process can be described as: by already installed camera, system front end obtains the video that video camera imports in real time, by local variance algorithm provided by the invention video is analyzed accurately and handled, extract the vehicle queue prospect of region of interest ROI, adopt the camera calibration algorithm to calculate the physical length of vehicle queue to the vehicle queue prospect, and the final vehicle queue length data that will calculate are passed to the system backstage, the traffic control system management and dispatching person assess the traffic route situation accordingly, and as a reference of traffic scheduling.(length of vehicle queue is a kind of as traffic parameter, can be used for estimating the traffic flow of road, also can be used as the control parameter of traffic lights.)
Vehicle queue length detects can be divided into three steps, image acquisition 1; Vehicle queue detects 2; Vehicle queue length calculates 3.Below this three step is specifically introduced:
One. as shown in Figure 2, in image acquisition 1, at first video camera is installed, adopt flash unit automatically image to be carried out light filling, the image that obtains is carried out the equalization operation earlier.And vehicle queue occurs when red light usually, so can judge traffic lights by the image processing method formula, when red light becomes green light, extracts a two field picture.
Two. vehicle queue detects 2.Employing is based on the vehicle queue detection algorithm of local variance, and specific algorithm is described below:
As shown in Figure 3:
Step (1): lane line is demarcated, for picture in its entirety, the vehicle that most interested is on the track, so in order to reduce calculated amount, improve arithmetic speed, at first lane line is demarcated, delimitation track zone ROI gets result and is " 1 ", for step (21), the follow-up processing respectively of step (2-2);
Step (21): in the hsv color space, tone H and saturation degree S have comprised colouring information, and brightness I is then irrelevant with color information, adopt tone component H to operate among the present invention; Result " 1 ", calculate each pixel in every 5*5 matrix-block tone component H square value average E{I (i, j)
2And the square value E of mean value (i, j)
2, again according to formula: D (i, j)=E{I (i, j)
2}-E (i, j)
2Calculate the variance D (i of the tone component H of each pixel, j), picture codomain scope after handling like this is bigger, variance D (i with all pixel tone component H, j) normalize in [01] scope, and adopting dual threshold method (this algorithm adopts T1=0.88 and T2=0.95) that picture is carried out binaryzation respectively, result is respectively " 3 " and " 2 ";
Again result " 2 " is corroded operation and obtain result " 4 ";
Simultaneously, step (2-2): the tone component H to result " 1 " in the step (1) handles, and adopt improved LBP algorithm to extract vehicle formation texture information and obtain result " 5 ",
Described improved LBP algorithm calculates according to following formula:
Step (3): step (2-1) result " 5 " and step (22) result " 4 " are carried out and operation, remove vehicle shadow, obtain result " 6 ";
Step (4): step (2-1) is handled back result " 3 " and step (3) handle afterwards that result " 6 " carries out or operates, adopt the morphology expansive working again, obtain vehicle formation prospect, and to prospect projection on the ROI center line, and definite head of the queue and tail of the queue place pixel.
Through behind the aforesaid operations, just can extract the vehicle queue prospect, and remove the influence of vehicle shadow.
Three. it is the vehicle queue length calculation flow chart that vehicle queue length calculates 3, Fig. 4, and the binary map after the vehicle formation is detected combines with the rule that obtains with the camera calibration algorithm and obtains the successive vehicles physical length.Wherein binary map is obtained by vehicle queue detection 2.Camera calibration, namely utilize shot by camera to image information reduce the geological information of object in the three dimensions.Therefore in computer vision, the homography on plane is defined as the projection mapping from a plane to another plane, and the mapping that is mapped on the video camera imaging instrument of the point on two dimensional surface can be represented with homography matrix H.
Be under the hypothesis of pin-hole model at video camera (3-1), according to four points demarcating on the picture and four points of objective plane, obtain homography matrix H;
(3-2) by homography matrix H, obtain the physical length that represents under the alive boundary of the center line epigraph two-dimensional image vegetarian refreshments coordinate system in the ROI, table that each pixel is corresponding with actual range among the present invention is called rule;
(3-3) tail of the queue and the team's head portrait vegetarian refreshments that arrives by the vehicle foreground detection obtained the length L1 of fleet according to the actual range of rule correspondence.
The error note that Fig. 5 produces for measuring vehicle formation method in the invention, the setting height(from bottom) H of video camera and overall height h can produce picture vision blind area, in order to improve the accuracy rate that vehicle queue length calculates, calculate vehicle queue length L=L1-l according to formula, wherein, L is vehicle commander's degree in the reality, d be video camera to the distance of stop line, the height of car is h, and triangular relationship h/H=l/ (d+L1) is arranged, calculate l=h (d+L1)/H, then final successive vehicles length L=L1 (1-h/H)-hd/H.
In sum, the present invention can be summarised as whole realization flow:
1) camera acquisition video.
2) judge lights state.
Gather a frame picture when 3) red light becomes green light.
4) camera calibration is carried out in the ROI zone, the computing scale chi.
5) carrying out vehicle queue according to the algorithm that proposes detects.
6) calculate vehicle queue length by prospect in the ROI and rule.
The present invention to state of the art innovative point:
1) proposes improved local variance method and carry out vehicle detection.
2) improved local variance and the improved LBP algorithm removal vehicle shadow that combines is proposed.
3) proposition reduces the method for the car queue length error of calculation.
Claims (1)
1. a method that detects the vehicle queue length of road junction is characterized in that, comprises the steps:
(1) adopt monocular-camera to obtain traffic video in real time at road junction;
(2) video that video camera is imported in real time carries out the virtual coil demarcation, handles knowledge according to image and judges lights state;
Image when gathering a frame vehicle queue when (3) red light becomes green light;
(4) region of interest ROI is carried out vehicle queue and detect, namely utilize improved local variance and improved LBP algorithm to combine and remove the shade of vehicle queue, improve the precision that vehicle queue detects, obtain vehicle queue prospect binary map in the ROI;
(5) camera calibration is carried out in the ROI zone, the computing scale chi.By homography matrix H, obtain the physical length that represents under the alive boundary of the center line epigraph two-dimensional image vegetarian refreshments coordinate system in the ROI, with each pixel and actual range one by one the table of corresponding acquisition be rule;
(6) integrating step (4) and both results of step (5), the tail of the queue that arrives by the vehicle foreground detection and team's head portrait vegetarian refreshments, and obtain the physical length L1 of road junction vehicle queue according to the actual range of rule correspondence.For setting height(from bottom) H and the overall height h that overcomes video camera can produce picture vision blind area, improve the accuracy rate that vehicle queue length calculates, the method that reply solves is:
Calculate vehicle queue length L=L1-1 according to formula, wherein, L is vehicle commander's degree in the reality, d be video camera to the distance of stop line, the height of car is h, and triangular relationship h/H=l/ (d+L1) is arranged, calculate l=h (d+L1)/H, then final successive vehicles length L=L1 (1-h/H)-hd/H;
The method of vehicle queue detection specific implementation is in the described ROI:
Step (1): lane line is demarcated, delimited out zone, track ROI, get result and be " 1 ", for step (21), the follow-up processing respectively of step (22);
Step (2-1): in the hsv color space, adopt tone component H to operate; Result " 1 " is calculated each pixel in every 5*5 matrix-block tone component H square value average E{I (i, j)
2And the square value E of mean value (i, j)
2, again according to formula: D (i, j)=E{I (i, j)
2}-E (i, j)
2Calculate the variance D (i of the tone component H of each pixel, j), picture codomain scope after handling like this is bigger, variance D (i with all pixel tone component H, j) normalize in [01] scope, and adopting the dual threshold method that picture is carried out binaryzation respectively, result is respectively " 3 " and " 2 ";
Again result " 2 " is corroded operation and obtain result " 4 ";
Simultaneously, step (22): select the tone component H of result " 1 " in the step (1) to handle, adopt improved LBP algorithm to extract vehicle formation texture information and obtain result " 5 ",
Described improved LBP algorithm calculates according to following formula:
Step (3): step (21) result " 5 " and step (2-2) result " 4 " are carried out and operation, remove vehicle shadow, obtain result " 6 ";
Step (4): step (21) is handled back result " 3 " and step (3) handle afterwards that result " 6 " carries out or operates, adopt the morphology expansive working again, obtain vehicle formation prospect, and to prospect projection on the ROI center line, and definite head of the queue and tail of the queue place pixel.
Through behind the aforesaid operations, extract the vehicle queue prospect, and removed the influence of vehicle shadow.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310134707.7A CN103268706B (en) | 2013-04-18 | 2013-04-18 | Method for detecting vehicle queue length based on local variance |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310134707.7A CN103268706B (en) | 2013-04-18 | 2013-04-18 | Method for detecting vehicle queue length based on local variance |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103268706A true CN103268706A (en) | 2013-08-28 |
| CN103268706B CN103268706B (en) | 2015-02-18 |
Family
ID=49012333
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310134707.7A Active CN103268706B (en) | 2013-04-18 | 2013-04-18 | Method for detecting vehicle queue length based on local variance |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103268706B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103606150A (en) * | 2013-11-15 | 2014-02-26 | 中国科学院东北地理与农业生态研究所 | A local variance quantification detection method for the size of a regular ground object spatial pattern |
| CN104835142A (en) * | 2015-03-10 | 2015-08-12 | 杭州电子科技大学 | Vehicle queuing length detection method based on texture features |
| CN107274673A (en) * | 2017-08-15 | 2017-10-20 | 苏州科技大学 | Vehicle queue length measuring method and measuring system based on amendment local variance |
| CN106128121B (en) * | 2016-07-05 | 2018-08-17 | 中国石油大学(华东) | Vehicle queue length fast algorithm of detecting based on Local Features Analysis |
| CN110164152A (en) * | 2019-07-03 | 2019-08-23 | 西安工业大学 | One kind being used for isolated traffic intersection traffic light control system |
| CN117437581A (en) * | 2023-12-20 | 2024-01-23 | 神思电子技术股份有限公司 | Motor vehicle congestion length calculation method based on image semantic segmentation and visual angle scaling |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002183880A (en) * | 2000-12-15 | 2002-06-28 | Toyo Commun Equip Co Ltd | Traffic situation providing method and device thereof |
| US20030190058A1 (en) * | 2002-04-04 | 2003-10-09 | Lg Industrial Systems Co., Ltd. | Apparatus and method for measuring queue length of vehicles |
| CN101469985A (en) * | 2007-12-26 | 2009-07-01 | 河海大学常州校区 | Single-frame image detection apparatus for vehicle queue length at road junction and its working method |
| CN101936730A (en) * | 2010-06-28 | 2011-01-05 | 汉王科技股份有限公司 | Vehicle queue length detection method and device |
| KR20110064814A (en) * | 2009-12-09 | 2011-06-15 | (주) 서돌 전자통신 | A traffic signal control system with aotomatic sensing of vehicle waiting for turn left signal by using image processing |
| CN102622897A (en) * | 2012-04-07 | 2012-08-01 | 山东大学 | Video-based dynamic vehicle queue length estimation method |
| CN102867414A (en) * | 2012-08-18 | 2013-01-09 | 湖南大学 | Vehicle queue length measurement method based on PTZ (Pan/Tilt/Zoom) camera fast calibration |
-
2013
- 2013-04-18 CN CN201310134707.7A patent/CN103268706B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002183880A (en) * | 2000-12-15 | 2002-06-28 | Toyo Commun Equip Co Ltd | Traffic situation providing method and device thereof |
| US20030190058A1 (en) * | 2002-04-04 | 2003-10-09 | Lg Industrial Systems Co., Ltd. | Apparatus and method for measuring queue length of vehicles |
| CN101469985A (en) * | 2007-12-26 | 2009-07-01 | 河海大学常州校区 | Single-frame image detection apparatus for vehicle queue length at road junction and its working method |
| KR20110064814A (en) * | 2009-12-09 | 2011-06-15 | (주) 서돌 전자통신 | A traffic signal control system with aotomatic sensing of vehicle waiting for turn left signal by using image processing |
| CN101936730A (en) * | 2010-06-28 | 2011-01-05 | 汉王科技股份有限公司 | Vehicle queue length detection method and device |
| CN102622897A (en) * | 2012-04-07 | 2012-08-01 | 山东大学 | Video-based dynamic vehicle queue length estimation method |
| CN102867414A (en) * | 2012-08-18 | 2013-01-09 | 湖南大学 | Vehicle queue length measurement method based on PTZ (Pan/Tilt/Zoom) camera fast calibration |
Non-Patent Citations (2)
| Title |
|---|
| 史忠科，等: "城市道路排队车辆检测方法", 《交通运输工程学报》 * |
| 郝灿，等: "基于改进型LBP特征的运动阴影去除算法", 《计算机系统应用》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103606150A (en) * | 2013-11-15 | 2014-02-26 | 中国科学院东北地理与农业生态研究所 | A local variance quantification detection method for the size of a regular ground object spatial pattern |
| CN103606150B (en) * | 2013-11-15 | 2016-03-30 | 中国科学院东北地理与农业生态研究所 | The method of a kind of local variance quantitative detection object space general layout size regularly |
| CN104835142A (en) * | 2015-03-10 | 2015-08-12 | 杭州电子科技大学 | Vehicle queuing length detection method based on texture features |
| CN104835142B (en) * | 2015-03-10 | 2017-11-07 | 杭州电子科技大学 | A kind of vehicle queue length detection method based on textural characteristics |
| CN106128121B (en) * | 2016-07-05 | 2018-08-17 | 中国石油大学(华东) | Vehicle queue length fast algorithm of detecting based on Local Features Analysis |
| CN107274673A (en) * | 2017-08-15 | 2017-10-20 | 苏州科技大学 | Vehicle queue length measuring method and measuring system based on amendment local variance |
| CN110164152A (en) * | 2019-07-03 | 2019-08-23 | 西安工业大学 | One kind being used for isolated traffic intersection traffic light control system |
| CN117437581A (en) * | 2023-12-20 | 2024-01-23 | 神思电子技术股份有限公司 | Motor vehicle congestion length calculation method based on image semantic segmentation and visual angle scaling |
| CN117437581B (en) * | 2023-12-20 | 2024-03-01 | 神思电子技术股份有限公司 | Motor vehicle congestion length calculation method based on image semantic segmentation and visual angle scaling |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103268706B (en) | 2015-02-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9704060B2 (en) | Method for detecting traffic violation | |
| CN103268706B (en) | Method for detecting vehicle queue length based on local variance | |
| CN106710240B (en) | The passing vehicle for merging multiple target radar and video information tracks speed-measuring method | |
| CN103310444B (en) | A kind of method of the monitoring people counting based on overhead camera head | |
| CN104392468B (en) | Based on the moving target detecting method for improving visual background extraction | |
| CN102496281B (en) | Vehicle red-light violation detection method based on combination of tracking and virtual loop | |
| CN102354457B (en) | General Hough transformation-based method for detecting position of traffic signal lamp | |
| CN105512720A (en) | Public transport vehicle passenger flow statistical method and system | |
| CN101567097B (en) | Bus passenger flow automatic counting method based on two-way parallactic space-time diagram and system thereof | |
| CN106128121B (en) | Vehicle queue length fast algorithm of detecting based on Local Features Analysis | |
| CN103871079A (en) | Vehicle tracking method based on machine learning and optical flow | |
| CN105184274B (en) | A kind of based on depth image acquisition passenger flow speed and the method for density parameter | |
| CN103902985B (en) | High-robustness real-time lane detection algorithm based on ROI | |
| CN106251695A (en) | Destination's parking stall intelligent recommendation system and method based on parking space state monitoring | |
| CN105513342A (en) | Video-tracking-based vehicle queuing length calculating method | |
| CN104183142A (en) | Traffic flow statistics method based on image visual processing technology | |
| CN103268470A (en) | Method for counting video objects in real time based on any scene | |
| CN104123734A (en) | Visible light and infrared detection result integration based moving target detection method | |
| CN103646254A (en) | High-density pedestrian detection method | |
| CN110009634A (en) | Vehicle count method in a kind of lane based on full convolutional network | |
| CN107274673B (en) | Vehicle queuing length measuring method and system based on corrected local variance | |
| CN104143077A (en) | Pedestrian target retrieving method and system based on images | |
| Cao et al. | Application of convolutional neural networks and image processing algorithms based on traffic video in vehicle taillight detection | |
| US10937202B2 (en) | Intensity data visualization | |
| CN106446832A (en) | Pedestrian real-time detection method based on video |
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 |