CN104464305A - Intelligent vehicle converse driving detecting device and method - Google Patents
Intelligent vehicle converse driving detecting device and method Download PDFInfo
- Publication number
- CN104464305A CN104464305A CN201410752578.2A CN201410752578A CN104464305A CN 104464305 A CN104464305 A CN 104464305A CN 201410752578 A CN201410752578 A CN 201410752578A CN 104464305 A CN104464305 A CN 104464305A
- Authority
- CN
- China
- Prior art keywords
- foreground
- prospect
- area
- highlight regions
- unit
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Abstract
The invention discloses an intelligent vehicle converse driving detecting device and method. The method comprises the steps that firstly, an image obtaining unit obtains video data in a monitoring area; secondly, a detection area setting unit is utilized for setting a detection area in a targeted mode; thirdly, a foreground acquiring unit is utilized for obtaining suspected foregrounds, and an effective foreground screening unit is used for removing part of false foregrounds; fourthly, a foreground classifying unit is utilized for carrying out classification about whether there is an automobile lamp area or not on the screened foreground, and a foreground historical data storage unit is used for carrying out respective storage; fifthly, a current frame foreground and a historical foreground are matched through a matching unit, and the historical foreground data are updated in real time through an updating unit; finally, a converse driving judgment unit is used for giving alarming information according to the foreground mass center motion direction. By means of the method for judging whether there is the automobile lamp or not in the foreground through distinguishing, the problem that the foreground is disorder and cannot be recognized due to the automobile lamp light rays is effectively solved.
Description
Technical field
The present invention relates to computer vision field, especially relate to the retrograde detection of the vehicle being applied to field of video monitoring.
Background technology
Drive in the wrong direction and refer to that vehicles or pedestrians not to keep to the side the behavior of advancing by the country one belongs to's requirement.It is that most endangers one of behavior of traffic safety that vehicle drives in the wrong direction.Along with the fast development of China's economy, the vehicle flowrate of each big city major trunk roads grows with each passing day.Retrograde event frequently occurs, as in two-way lane, track, side blocks up a little, have quite a few vehicle drive in the wrong direction pass through the district that blocks up.This not only can bring new blocking up also can bring hidden peril of accident, and retrograde vehicle in tunnel is more dangerous.The traffic hazard caused because driving in the wrong direction happens occasionally.
Therefore, a kind of effective detection means of driving in the wrong direction for vehicle is fast needed.Traditional vehicle detection means of driving in the wrong direction mainly contains Data mining, and microwave detects.And Data mining installs or safeguard all to need road closure to suspend traffic, and be difficult to realize other functions; The setting parameter intuitive that microwave detects is poor, and maintenance cost is higher.
Summary of the invention
The object of the invention is to propose vehicle to drive in the wrong direction intelligent detection device and method.The present invention can provide warning when vehicle drives in the wrong direction fast and effectively by video analysis.By prospect being classified with or without car light, effectively solve robustness problem.And install simple and easy, easy to maintenance, extendable functions is many.
In order to achieve the above object, the present invention proposes vehicle and to drive in the wrong direction intelligent detection device and method:
The device that the present invention proposes comprises as lower unit:
Image acquisition unit, by the video data of camera Real-time Collection appointed area;
Surveyed area setting unit, by setting surveyed area targetedly, gets rid of the interference of non-surveyed area;
Prospect asks for unit, utilizes the method for background modeling and foreground detection to ask for prospect in video image;
Effective prospect discriminator unit, utilization presets relevant restrictive condition and gets rid of some ineligible prospects;
Foreground classification unit, is classified prospect with or without highlight regions by differentiation prospect;
Prospect history data store unit, stores respectively by sorted prospect unit, as the historical data of next frame coupling;
Matching unit, mates the foreground data of present frame with the historical data of preservation;
Updating block, upgrades the history foreground data present frame foreground data that the match is successful;
Drive in the wrong direction judging unit, judges whether to provide warning by the center of mass motion direction of prospect each in detection history data.
The method that the present invention proposes comprises the steps:
First step, utilizes camera to obtain the video data of monitoring road;
Second step, the position of guarded region needed for the different set of foundation monitoring site surrounding and size;
Third step, asks for the vehicle foreground of opening car light;
4th step, asks for the vehicle foreground of lamp of not driving;
5th step, gets rid of some false prospects and undesirable prospect by setting correlated condition;
Whether the 6th step, exist highlight regions to define different foreground type according in prospect;
7th step, preserves the relevant foreground data with highlight regions, it can be used as historical data;
8th step, preserves the relevant foreground data not having highlight regions; It can be used as historical data;
9th step, does matching treatment by present frame foreground data with corresponding history foreground data;
Tenth step, upgrades the history foreground data present frame foreground data that the match is successful;
11 step, by judging that prospect center of mass motion direction and matching times provide warning message.
The invention has the beneficial effects as follows: during early morning, each vehicle car light switch differs at dusk, retrograde vehicle opens distance light often, now car light light causes very large difficulty for the foreground detection of general algorithm, whether the present invention opens car light by distinguishing vehicle, carry out different prospect acquiring methods, effectively solve this problem, greatly improve the accuracy rate of detection.
Accompanying drawing explanation
The present invention has six, accompanying drawing:
Fig. 1 shows the schematic block diagram of the intelligent detection device that to drive in the wrong direction according to vehicle of the present invention;
Fig. 2 shows the schematic block diagram asking for unit according to vehicle of the present invention prospect in intelligent detection device of driving in the wrong direction;
Fig. 3 shows and to drive in the wrong direction the schematic block diagram of judging unit of driving in the wrong direction in intelligent arrangement for detecting according to vehicle of the present invention;
Fig. 4 shows the overall flow figure of the intelligent detecting method that to drive in the wrong direction according to vehicle of the present invention;
Fig. 5 shows according to the third step process flow diagram based on video smoke detection method of the present invention;
Fig. 6 shows according to the 11 flow chart of steps based on video smoke detection method of the present invention.
Embodiment
With reference to the accompanying drawings and the present invention is described in detail in conjunction with instantiation.The YUV coloured image of 352 × 288 pixel sizes that embodiment adopts.Be to be noted that described example is only for the ease of the understanding of the present invention, therefore do not limit protection scope of the present invention.
Fig. 1 shows the schematic block diagram of the intelligent detection device that to drive in the wrong direction according to vehicle of the present invention.The each unit of the following stated device all can realize separately through microprocessor, but a processor with high-performance CPU also can be utilized to realize for cost-saving.As shown in Figure 1, comprise according to the vehicle of the present invention intelligent detection device that drives in the wrong direction:
Part I, image acquisition unit 101, obtains the video data of guarded region by camera.
Part II, surveyed area setting unit 102, according to the image information that image acquisition unit 101 gathers, the region that will detect is set in the picture, be not arranged to the part of surveyed area, do not carry out driving in the wrong direction and detect, calculated amount can not only be reduced like this and unnecessary interference and wrong report can be got rid of.
Part III, prospect asks for unit 103, adopts the mode of background modeling and foreground detection to carry out foreground extraction to the surveyed area that surveyed area setting unit 102 is arranged; The prospect of wherein figure 2 show asks for the schematic block diagram of unit 103, comprising:
Background modeling unit 201, utilizes multiple image to set up main background by Gauss's modeling method;
Foreground detection unit 202, is compared by present frame and background model, asks for doubtful foreground area;
Highlight regions judging unit 203, judges whether the doubtful foreground area that foreground detection unit 202 extracts exists highlight regions, and different according to the result judged, unit carries out different disposal afterwards;
Parameter calculation unit 204, according to the result of highlight regions judging unit 203, process accordingly, if there is not highlight regions in foreground area, then calculate the area of prospect, girth, filling rate, the ratio of width to height, barycenter, if foreground area exists highlight regions, then calculate outside above-mentioned prospect parameter, also will the ratio of whole foreground area shared by the barycenter of highlight regions, highlight regions area in calculating prospect.
Part IV, effective prospect discriminator unit 104, get rid of some ineligible false prospects by the filling rate of foregrounding, the ratio of width to height, area, average gray, also will judge that shared by highlight regions place foreground location and highlight regions area, foreground area ratio gets rid of false prospect further for the prospect with highlight regions in addition.
Part V, foreground classification unit 105, by judging whether foreground area exists highlight regions and the prospect that effective prospect discriminator unit 104 retains is carried out classification process.
Part VI, prospect history data store unit 106, is used for storing the sorted foreground data of foreground classification unit 105.
Part VII, matching unit 107, carries out matching operation by present frame foreground data and history foreground data.
Part VIII, updating block 108, matching result according to matching unit 107 is different, carries out different process, if present frame foreground data matches with corresponding history foreground data to historical data, then corresponding history foreground data matching times adds 1, by present frame foreground data more new historical foreground data, if result is not mated, then corresponding history foreground data loss number of times adds 1, lose number of times when history foreground data and reach setting threshold value, process is emptied to it.
Part IX, drive in the wrong direction judging unit 109, and specifically composition is as shown in Figure 3:
Coupling foreground area extraction unit 301, is used for being extracted in history prospect nearest with present frame prospect on image;
Highlight regions judging unit 302, judges whether the present frame prospect that coupling foreground area extraction unit 301 extracts and history prospect exist highlight regions;
Center of mass motion walking direction unit 303, different according to the result that highlight regions judging unit 302 judges, adopt different judgment modes, if prospect exists highlight regions, then judge highlight regions center of mass motion direction, if prospect does not exist highlight regions, then judge overall foreground area center of mass motion direction, record present frame centroid position is also made comparisons with history centroid position, if center of mass motion direction is contrary with direction initialization, then it is retrograde for recording present frame;
Judge alarm unit 304, the retrograde frame number added up according to center of mass motion walking direction unit 303 compared with the threshold value of setting, thus judges whether to provide warning message.
Fig. 4 shows the process flow diagram of the intelligent detecting method that to drive in the wrong direction according to vehicle of the present invention, due to the difference of actual scene and the erection parameter of camera different, the threshold value in following can different, and best threshold value could be determined after need testing according to reality.Therefore, no longer enumerate actual data to be here described.As shown in Figure 4, overall flow of the present invention is divided into 11 steps:
Step 401, video image obtains, and is obtained the video data of guarded region by common camera.
Step 402, arrange surveyed area, the specific aim of carrying out surveyed area according to the video image obtained is arranged, and is not arranged to the part of surveyed area, do not carry out driving in the wrong direction and detect, calculated amount can not only be reduced like this and unnecessary interference and wrong report can be got rid of.
Step 403, ask for band car light prospect, during dark, as night, at the cloudy day, under the conditions such as haze, vehicle major part opens car light, if now still detect overall vehicle prospect, effect is very undesirable, therefore the present invention adopts the detection difficult problem asked for when car light prospect overcomes dark greatly as the method for Rule of judgment, step 403 idiographic flow as shown in Figure 5:
Step 501 adopts the mode of carrying out Gauss's modeling to continuous multiple frames image to set up and detects main background, the background that step 502 utilizes step 501 to be successfully established asks for prospect to surveyed area, in step 503 determining step 502, whether required foreground area exists highlight regions, step 504 carries out different operations according to the testing result difference of step 503, if step 503 does not detect that prospect exists highlight regions, then step 504 calculates the area of foreground area, girth, filling rate, the ratio of width to height, barycenter, if step 503 detects in prospect to there is highlight regions, then except calculating above-mentioned parameter, also to calculate the ratio of the prospect total area shared by the centroid position of highlight regions and highlight regions area.
The Gaussian Background modeling principle that the present invention takes is: modeling point adopts the pixel of 2 × 2 sizes, each modeling point sets up three models, this modeling point Y deposited by each model, U, V information, according to the match is successful, number of times arranges three models corresponding to modeling point from big to small, image first frame is directly assigned to the first model in background model, each frame is according to the Y of present frame modeling point afterwards, U, V data carry out matching operation with the corresponding model of this modeling point respectively, coupling is judged whether by setting threshold value T1, namely the yuv data of present frame modeling point differs within threshold value T1 with background model yuv data, then think and this Model Matching, and the matching times of this model is added 1, if current modeling point does not all mate with three models, replaced the 3rd model, namely the model that matching times is minimum, when in three models that modeling point is corresponding, any one matching times reaches threshold value T2, then this modeling point is considered as modeling success, now by maximum for matching times in corresponding for modeling point model and modeling successful model assignment gives main background, by adjustment threshold value T1, the sensitivity of foreground detection can be adjusted, by adjustment threshold value T2, the successful speed of modeling can be adjusted, the optimal threshold of T1 and T2 need carry out respective settings according to the vehicle flowrate of actual traffic road.
Step 404, asks for and is not with car light prospect, described mode and step 403 similar, still according to the mode of background modeling and foreground detection, ask for the doubtful prospect in guarded region, and calculate area, girth, barycenter, filling rate, the ratio of width to height of required prospect.
Step 405, screen effective foreground information, the false prospect not meeting pre-conditioned threshold value is got rid of, for the judgement having the prospect of highlight regions also will carry out overall foreground area ratio shared by highlight regions size and highlight regions according to the area set, filling rate, the ratio of width to height threshold value.
Vehicle image information reflected in video, the area of its vehicle foreground, filling rate, the ratio of width to height size always change in a relatively-stationary scope, therefore get rid of the excessive or too small false prospect of some areas, filling rate and the ratio of width to height by arranging suitable threshold value; Actual threshold value is different with angle and different according to the height of lens focus and camera erection, need could obtain optimal parameter threshold value according to field adjustable.
Step 406, foreground type defines, by judging whether required prospect exists highlight regions and carry out foreground type differentiation, the Different Results that storage operation afterwards and matching operation can be judged according to current procedures and carry out different process.
Step 407, preserves the advance data with highlight regions, comprises ratio and the highlight regions centroid position of overall foreground area area shared by the area of prospect, girth, filling rate, barycenter, the ratio of width to height, highlight regions area according to step 406 result of determination.
Step 408, preserves the foreground data without highlight regions, comprises the area of prospect, girth, filling rate, barycenter, the ratio of width to height according to step 406 result of determination.
Step 409, present frame prospect is mated with history prospect, be chosen at history foreground data nearest with present frame prospect on image to mate, matching operation comprises area, filling rate, the ratio of width to height of comparing the two, if differ within setting threshold value, then think and the two coupling also will compare overall proportion of foreground shared by the two highlight regions in addition for the prospect with highlight regions whether identical.
Step 410, history foreground data upgrades, matching result according to step 409 carries out respective handling to history foreground data, for the history prospect that the match is successful, by present frame foreground data more new historical foreground data matching times is added 1, for the history foreground data that it fails to match, its lost frames number of times is added 1, add up the lost frames number of times of each history foreground data, when reaching setting threshold value, show that this history foreground data loses number of times too much, carry out emptying process, and for there is no the present frame prospect that the match is successful, be stored as new history foreground data.
Step 411, judge to report to the police, the coupling frame number added up according to step 410 and every frame centroid position of record carry out last retrograde alarm decision, idiographic flow as shown in Figure 6:
Step 601 is by prospect centroid position relation, extract the history prospect nearest with present frame prospect, whether the present frame prospect that step 602 determining step 601 extracts and history prospect exist highlight regions, the result that step 603 judges according to step 602 is different and take different operations, if foreground area exists high bright spot, choose highlight regions barycenter as Rule of judgment, if foreground area does not exist highlight regions, choose overall prospect barycenter as Rule of judgment, judge present frame prospect barycenter and history prospect centroid position relation, if center of mass motion direction and direction initialization are contrary, the corresponding history foreground data record frame number that drives in the wrong direction adds 1, step 604 compares according to the retrograde frame number of times of history foreground data record and setting threshold value, provide warning message.
The above; be only the embodiment in the present invention, but protection scope of the present invention is not limited thereto, any researchist being familiar with this technology is in the technical scope disclosed by the present invention; according to the multiple change that actual conditions are made, all should be encompassed within protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with the protection domain of claims.
Claims (6)
1. vehicle drives in the wrong direction intelligent detection device and method, and it is characterized in that, this device comprises:
Part I, image acquisition unit, by the video data of camera Real-time Collection appointed area;
Part II, surveyed area setting unit, by setting surveyed area targetedly, gets rid of the interference of non-surveyed area;
Part III, prospect asks for unit, utilizes the method for background modeling and foreground detection to ask for prospect in video image;
Part IV, effective prospect discriminator unit, utilization presets relevant restrictive condition and gets rid of some ineligible prospects;
Part V, foreground classification unit, is classified prospect with or without highlight regions by differentiation prospect;
Part VI, prospect history data store unit, stores respectively by sorted prospect unit, as the historical data of next frame coupling;
Part VII, matching unit, mates the foreground data of present frame with the historical data of preservation;
Part VIII, updating block, upgrades the history foreground data present frame foreground data that the match is successful;
Part IX, drive in the wrong direction judging unit, judges whether to provide warning by the center of mass motion direction of prospect each in detection history data.
2. according to device according to claim 1, it is characterized in that, prospect is asked for unit and is comprised:
A background modeling unit, utilizes multiple image to set up main background by Gauss's modeling method; B foreground detection unit, is compared by present frame and background model, asks for doubtful foreground area; C highlight regions judging unit, judges whether the doubtful foreground area that foreground detection unit extracts exists highlight regions, processes accordingly according to the result judged; D parameter calculation unit, result according to highlight regions judging unit carries out respective handling, if there is not highlight regions in foreground area, then calculate the area of prospect, girth, filling rate, the ratio of width to height, barycenter, if there is highlight regions in foreground area, then calculate outside above-mentioned prospect parameter, also will the ratio of whole foreground area shared by the barycenter of highlight regions, highlight regions area in calculating prospect.
3. according to device according to claim 1, it is characterized in that, retrograde judging unit comprises:
Coupling foreground area extraction unit, is used for being extracted in history prospect nearest with present frame prospect on image; Highlight regions judging unit, judges whether the present frame that coupling foreground area extraction unit extracts and history prospect exist highlight regions; Center of mass motion walking direction unit, different according to the result that highlight regions judging unit judges, adopt different judgment modes, if prospect exists highlight regions, then judge highlight regions center of mass motion direction, if prospect does not exist highlight regions, then judge overall foreground area center of mass motion direction, record present frame foreground area centroid position is also made comparisons with corresponding history foreground area centroid position, if center of mass motion direction is contrary with direction initialization, then it is retrograde for recording present frame; Judge alarm unit, the retrograde frame number added up according to center of mass motion walking direction unit compared with the threshold value of setting, thus judges whether to provide warning message.
4. vehicle drives in the wrong direction intelligent detection device and method, and it is characterized in that, the method step comprises:
First step, video image obtains, and is obtained the video data of guarded region by common camera;
Second step, arranges surveyed area, and the specific aim of carrying out surveyed area according to the video image obtained is arranged;
Third step, ask for band car light prospect, foreground detection is carried out according to background model, calculate area, girth, filling rate, the aspect ratio information of foreground area, if there is highlight regions in foreground area, then also need the ratio calculating the prospect total area shared by the centroid position of highlight regions and highlight regions area
4th step, asks for and is not with car light prospect, carries out foreground detection according to background model, calculates area, girth, filling rate, the aspect ratio information of foreground area;
5th step, screen effective foreground information, pre-conditioned false prospect is not met, for the judgement having the prospect of highlight regions also will carry out overall foreground area ratio shared by highlight regions size and highlight regions according to the area set, girth, filling rate, the eliminating of the ratio of width to height threshold value;
6th step, foreground type defines, by judging whether required prospect exists the high region that connects and carry out foreground type differentiation;
7th step, preserves the advance data with highlight regions;
8th step, preserves the foreground data without highlight regions;
9th step, prospect is mated, be chosen at history foreground data nearest with present frame prospect on image to mate, matching operation comprises area, filling rate, the ratio of width to height of comparing the two, if differ within setting threshold value, then think and the two coupling also will compare overall proportion of foreground shared by the two highlight regions in addition for the prospect with highlight regions whether identical;
Tenth step, history foreground data upgrades, according to matching result, respective handling is carried out to history foreground data, for the history prospect that the match is successful, by present frame foreground data more new historical foreground data matching times is added 1, for the history foreground data that it fails to match, its lost frames number of times is added 1, add up the lost frames number of times of each history foreground data, when reaching setting threshold value, showing that this history foreground data loses number of times too much, carrying out emptying process, and for there is no the present frame prospect that the match is successful, be stored as new history foreground data;
11 step, judges to report to the police, and carries out last retrograde alarm decision according to the coupling frame number of statistics and every frame centroid position of record.
5. in accordance with the method for claim 4, it is characterized in that, third step comprises following sub-step:
A employing sets up the main background of detection to the mode that continuous multiple frames image carries out Gauss's modeling, B utilizes the background be successfully established to ask for prospect to surveyed area, C detects required foreground area and whether there is highlight regions, D carries out different operations according to testing result difference, if there is highlight regions in the prospect of not detecting, then calculate the area of foreground area, girth, filling rate, the ratio of width to height, barycenter, if detect in prospect to there is highlight regions, then except calculating above-mentioned parameter, also to calculate the ratio of the prospect total area shared by the centroid position of highlight regions and highlight regions area.
6. in accordance with the method for claim 4, it is characterized in that, the 11 step comprises following sub-step:
E is by prospect centroid position relation, extract the history prospect nearest with present frame prospect, F judges whether the present frame prospect extracted and history prospect exist highlight regions, G takes different operations according to the result difference judged, if foreground area exists highlight regions, choose highlight regions barycenter as Rule of judgment, if foreground area does not exist highlight regions, choose overall prospect barycenter as Rule of judgment, judge present frame prospect barycenter and history prospect centroid position relation, if center of mass motion direction and direction initialization are contrary, the corresponding history foreground data record frame number that drives in the wrong direction adds 1, H compares according to the retrograde frame number of times of history foreground data record and setting threshold value, provide warning message.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410752578.2A CN104464305A (en) | 2014-12-11 | 2014-12-11 | Intelligent vehicle converse driving detecting device and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410752578.2A CN104464305A (en) | 2014-12-11 | 2014-12-11 | Intelligent vehicle converse driving detecting device and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104464305A true CN104464305A (en) | 2015-03-25 |
Family
ID=52910280
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410752578.2A Pending CN104464305A (en) | 2014-12-11 | 2014-12-11 | Intelligent vehicle converse driving detecting device and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104464305A (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104809887A (en) * | 2015-04-23 | 2015-07-29 | 杭州中威电子股份有限公司 | Detecting method for vehicle converse running on expressway and automatic alarm device |
| CN105120239A (en) * | 2015-09-18 | 2015-12-02 | 天津艾思科尔科技有限公司 | Detector with anti-collision protective structure |
| CN105354529A (en) * | 2015-08-04 | 2016-02-24 | 北京时代云英科技有限公司 | Vehicle converse running detection method and apparatus |
| CN105957312A (en) * | 2016-06-06 | 2016-09-21 | 重庆微标科技股份有限公司 | Prompting method and system for abnormal clearance of vehicle |
| CN106504542A (en) * | 2016-09-30 | 2017-03-15 | 北京中星微电子有限公司 | Speed intelligent monitoring method and system |
| CN106846297A (en) * | 2016-12-21 | 2017-06-13 | 深圳市镭神智能系统有限公司 | Pedestrian's flow quantity detecting system and method based on laser radar |
| WO2018068317A1 (en) * | 2016-10-14 | 2018-04-19 | 富士通株式会社 | Vehicle wrong-way detection apparatus and method, and electronic device |
| CN108027874A (en) * | 2015-07-30 | 2018-05-11 | 灯塔人工智能公司 | Use the security system based on computer vision of depth camera |
| CN108204808A (en) * | 2016-12-16 | 2018-06-26 | 卡西欧计算机株式会社 | Map matching means, map-matching method and recording medium |
| CN109313852A (en) * | 2016-06-07 | 2019-02-05 | 罗伯特·博世有限公司 | Method, apparatus and system for driver's identification of driving in the wrong direction |
| CN110428660A (en) * | 2019-08-12 | 2019-11-08 | 中车株洲电力机车研究所有限公司 | A kind of identification of self- steering vehicle line traffic direction and control method, device, system and computer readable storage medium |
| CN110689734A (en) * | 2019-09-24 | 2020-01-14 | 成都通甲优博科技有限责任公司 | Vehicle running condition identification method and device and electronic equipment |
| CN112466003A (en) * | 2019-09-06 | 2021-03-09 | 顺丰科技有限公司 | Vehicle state detection method, device, server and storage medium |
| CN113327414A (en) * | 2020-02-28 | 2021-08-31 | 深圳市丰驰顺行信息技术有限公司 | Vehicle reverse running detection method and device, computer equipment and storage medium |
| WO2022134387A1 (en) * | 2020-12-21 | 2022-06-30 | 深圳市商汤科技有限公司 | Vehicle wrong-way travel detection method, apparatus, device, computer-readable storage medium, and computer program product |
-
2014
- 2014-12-11 CN CN201410752578.2A patent/CN104464305A/en active Pending
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104809887A (en) * | 2015-04-23 | 2015-07-29 | 杭州中威电子股份有限公司 | Detecting method for vehicle converse running on expressway and automatic alarm device |
| CN108027874A (en) * | 2015-07-30 | 2018-05-11 | 灯塔人工智能公司 | Use the security system based on computer vision of depth camera |
| CN108027874B (en) * | 2015-07-30 | 2022-11-18 | 苹果公司 | Computer vision based security system using depth camera |
| CN105354529A (en) * | 2015-08-04 | 2016-02-24 | 北京时代云英科技有限公司 | Vehicle converse running detection method and apparatus |
| CN105120239A (en) * | 2015-09-18 | 2015-12-02 | 天津艾思科尔科技有限公司 | Detector with anti-collision protective structure |
| CN105957312A (en) * | 2016-06-06 | 2016-09-21 | 重庆微标科技股份有限公司 | Prompting method and system for abnormal clearance of vehicle |
| CN109313852A (en) * | 2016-06-07 | 2019-02-05 | 罗伯特·博世有限公司 | Method, apparatus and system for driver's identification of driving in the wrong direction |
| CN106504542A (en) * | 2016-09-30 | 2017-03-15 | 北京中星微电子有限公司 | Speed intelligent monitoring method and system |
| WO2018068317A1 (en) * | 2016-10-14 | 2018-04-19 | 富士通株式会社 | Vehicle wrong-way detection apparatus and method, and electronic device |
| CN109661667A (en) * | 2016-10-14 | 2019-04-19 | 富士通株式会社 | The retrograde detection device of vehicle and method, electronic equipment |
| JP2019532431A (en) * | 2016-10-14 | 2019-11-07 | 富士通株式会社 | Vehicular retrograde detection device, detection method, and electronic apparatus |
| CN108204808A (en) * | 2016-12-16 | 2018-06-26 | 卡西欧计算机株式会社 | Map matching means, map-matching method and recording medium |
| CN108204808B (en) * | 2016-12-16 | 2021-11-02 | 卡西欧计算机株式会社 | Map matching device, map matching method, and computer program product |
| CN106846297A (en) * | 2016-12-21 | 2017-06-13 | 深圳市镭神智能系统有限公司 | Pedestrian's flow quantity detecting system and method based on laser radar |
| CN110428660A (en) * | 2019-08-12 | 2019-11-08 | 中车株洲电力机车研究所有限公司 | A kind of identification of self- steering vehicle line traffic direction and control method, device, system and computer readable storage medium |
| CN112466003A (en) * | 2019-09-06 | 2021-03-09 | 顺丰科技有限公司 | Vehicle state detection method, device, server and storage medium |
| CN112466003B (en) * | 2019-09-06 | 2023-11-28 | 顺丰科技有限公司 | Vehicle state detection method, device, computer equipment and storage medium |
| CN110689734A (en) * | 2019-09-24 | 2020-01-14 | 成都通甲优博科技有限责任公司 | Vehicle running condition identification method and device and electronic equipment |
| CN113327414A (en) * | 2020-02-28 | 2021-08-31 | 深圳市丰驰顺行信息技术有限公司 | Vehicle reverse running detection method and device, computer equipment and storage medium |
| WO2022134387A1 (en) * | 2020-12-21 | 2022-06-30 | 深圳市商汤科技有限公司 | Vehicle wrong-way travel detection method, apparatus, device, computer-readable storage medium, and computer program product |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104464305A (en) | Intelligent vehicle converse driving detecting device and method | |
| CN103914688B (en) | A kind of urban road differentiating obstacle | |
| CN102385803B (en) | All-weather urban vehicle tracking and counting method based on video monitoring | |
| CN103400111A (en) | Method for detecting fire accident on expressway or in tunnel based on video detection technology | |
| KR101095528B1 (en) | An outomatic sensing system for traffic accident and method thereof | |
| US9704060B2 (en) | Method for detecting traffic violation | |
| CN105512720A (en) | Public transport vehicle passenger flow statistical method and system | |
| CN102521983B (en) | Vehicle violation detection system based on high definition video technology and method thereof | |
| CN103714363B (en) | A kind of motor vehicle exhaust smoke video identification system | |
| CN104657735B (en) | Method for detecting lane lines, system, lane departure warning method and system | |
| CN105913685A (en) | Video surveillance-based carport recognition and intelligent guide method | |
| CN110862033B (en) | Intelligent early warning detection method applied to inclined shaft winch of coal mine | |
| CN106652551A (en) | Parking stall detection method and device | |
| CN106652465A (en) | Method and system for identifying abnormal driving behavior on road | |
| CN201773466U (en) | Video monitoring and pre-warning device for detecting, tracking and identifying object detention/stealing event | |
| CN102819764A (en) | Method for counting pedestrian flow from multiple views under complex scene of traffic junction | |
| CN105744232A (en) | Method for preventing power transmission line from being externally broken through video based on behaviour analysis technology | |
| CN107478590A (en) | A kind of method of combination motor vehicle intelligent vision identification and remote exhaust emission detection | |
| CN102142194B (en) | Video detection method and system | |
| CN103021059A (en) | Video-monitoring-based public transport passenger flow counting method | |
| CN101656023A (en) | Management method of indoor car park in video monitor mode | |
| CN201307337Y (en) | Automatic alarming and monitoring device for traffic-lights nonobservance of pedestrian | |
| CN104504908A (en) | Monitoring method and system for illegal parking | |
| CN102897086A (en) | Back car driving information detecting and prompting method and system | |
| CN107590834A (en) | A kind of road traffic accident video detecting method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20150325 |