CN106295459A - Based on machine vision and the vehicle detection of cascade classifier and method for early warning - Google Patents
Based on machine vision and the vehicle detection of cascade classifier and method for early warning Download PDFInfo
- Publication number
- CN106295459A CN106295459A CN201510236672.7A CN201510236672A CN106295459A CN 106295459 A CN106295459 A CN 106295459A CN 201510236672 A CN201510236672 A CN 201510236672A CN 106295459 A CN106295459 A CN 106295459A
- Authority
- CN
- China
- Prior art keywords
- frame
- vehicle
- target
- width
- early warning
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/58—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
- G06V20/584—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads of vehicle lights or traffic lights
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/16—Anti-collision systems
- G08G1/166—Anti-collision systems for active traffic, e.g. moving vehicles, pedestrians, bikes
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Traffic Control Systems (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The present invention provides a kind of based on machine vision and the vehicle detection of cascade classifier with method for early warning, comprises the steps: to utilize the auto graph through the most known vehicle of editing to carry out the cascade classifier training of positive negative sample；The mathematical model identifying width of frame with reality distance is set up by a large amount of measured datas；Real time imaging obtains；Image semantic classification, confines area-of-interest；Area-of-interest is carried out target recognition；Recognition result is carried out data analysis, thus screens and follow the tracks of target；It is identified the process of frame moving average filtering to following the tracks of target；Calculate the actual range identifying target, calculated the translational speed of vehicle by frame difference time and actual range difference, obtain with this it may happen that time of colliding show early warning information.The method of the present invention, by utilizing positive and negative sample training cascade classifier, real time imaging acquisition, Image semantic classification, target recognition, target following and the process such as screening, target alert, is effectively increased in actual environment the recognition efficiency to mobile target and enhances under complex environment detection and the trace ability to moving target.
Description
Technical field
The present invention relates to spacing intelligent measure and computing technique field, specifically, relate to a kind of based on machine vision and the vehicle detection of cascade classifier with method for early warning.
Background technology
At present, driver is when driving, due to reasons such as driving fatigues, usually need a kind of instrument effectively carrying out anti-collision warning during dispatch driving, carry out anti-collision warning, first have to solve real time distance problem, the most conventional distance measuring method has four kinds: ultrasonic ranging, millimetre-wave radar is found range, laser ranging, and camera system is found range.
Ultrasonic ranging refers to the method for distance utilizing the time difference launching and receiving ultrasound wave to calculate target.Ultrasonic range finder has principle simple, and easy to make, cost is than relatively low advantage.But affected by environment relatively big, and it is unsuitable for distance range finding.
Radar utilizes target to reflection of electromagnetic wave to find target and to measure its position.As long-range sensor, radar range finding has detection performance stable, and being susceptible to characteristics of objects affects, the advantage of good environmental adaptability, but relatively costly.
Laser range finder determines distance according to the laser beam propagation time, has that the measurement time is short, range big, precision advantages of higher, but higher to stability, reliability requirement, and volume is also restrained, and cost is the highest simultaneously.
Camera system range finding generally uses biocular systems, utilizes two video cameras simultaneously to same Scenery Imaging, by two width image analysis processing, it may be determined that the three-dimensional coordinate of object.Biocular systems has the good characteristics such as certainty of measurement is high, size is little, low in energy consumption, dynamic range big, light accurate measurement, but its price is higher, and image taking speed is slow, and is restricted by hardware.
It is poor to there is precision in above several method, relatively costly, function singleness, need to install expensive hardware equipment additional, image taking speed is slow, the shortcoming such as bigger affected by environment, is not particularly suited for remote real time speed measuring and real-time collision early warning, therefore, current driver is in the urgent need to a kind of instrument carrying out anti-collision warning during dispatch driving easy to use, effective.
Summary of the invention
The present invention be a kind of based on machine vision technique by the cascade classifier method to vertically moving automobile detecting following real-time early warning, overcome deficiency of the prior art, improve in actual environment the recognition efficiency of mobile target and enhance under complex environment detection and the trace ability to moving target.
The technical scheme is that a kind of based on machine vision and the vehicle detection of cascade classifier with method for early warning, comprise the steps:
A. the auto graph through the most known vehicle of editing is utilized to carry out the cascade classifier training of positive negative sample；
B. the mathematical model identifying width of frame with reality distance is set up by a large amount of measured datas；
C. real time imaging obtains；
D. Image semantic classification, confines area-of-interest；
E. area-of-interest is carried out target recognition；
F. recognition result is carried out data analysis, thus screens and follow the tracks of target；
G. tracking target is identified frame moving average filtering to process；
H. calculate the actual range identifying target, calculated the translational speed of vehicle by frame difference time and actual range difference, obtain with this it may happen that time of colliding show early warning information.
Wherein, described grader LBP based on vehicle tail features training obtains.
In step B, mathematical model is set up in the following manner: by a large amount of measured datas, obtains in entity and image width and the dependency relation of distance and carries out fitting function error analysis, determining range finding mathematical model.
In step D, carry out by following procedure:
A. original image is converted into OPENCV accessible Mat object；
B. to mat object scaling and gray proces；
C. gray level image is carried out gray-level histogram equalization；
D. use mask shade image that regions of non-interest realizes image to cover, thus reduce area-of-interest to reduce the complex environment interference to grader identification.
Step E includes that goal ordering and vehicle front detect process, specific as follows:
Goal ordering refers to be ranked up to the distance of oneself according to target, and the distance of target is corresponding with the y-coordinate value identifying vehicle in machine vision, and distance is the most remote, and it identifies that the y value of frame is the biggest, is identified that by contrast the y-coordinate value of frame is ranked up；
Vehicle front detection refers to the vehicle that oneself may be threatened by front in longitudinal driving process, here by point at the bottom of as triangle two of the transverse width in oneself vehicle width correspondence video known, road infinite point in video is considered as vertex of a triangle, connect three points triangle, when identifying that frame is considered as front vehicles when colliding with this detection triangle, the front vehicle to oneself being likely to result in threat is filtered out with this, simultaneously according to identifying the sequence of frame y-coordinate value, find apart from oneself nearest vehicle.
In step F, contrasted by the identification frame coordinate of previous frame with next frame, use Euclidean distance to calculate the distance of two coordinates, when distance is considered as same car in threshold range, take that threshold value is vehicle identification width of frame here 1/2nd；Use moving average filtering processing method to identifying that width is smoothed, to the vehicle tracked by the identification width calculation meansigma methods of its front four frames；
Wherein, pretreatment is carried out before to identification frame smoothing processing, detailed process is as follows: by the identification width of frame contrast identifying width of frame and previous frame of present frame, when two frame identification width of frame differences exceed the 0.1 of previous frame width, this frame identification width of frame is adjusted to the meansigma methods of present frame identification width of frame and previous frame identification width of frame.
Provided by the present invention based on machine vision and the vehicle detection of cascade classifier with method for early warning, have the advantage that
First: the method for the present invention, by utilizing positive and negative sample training cascade classifier, real time imaging acquisition, Image semantic classification, target recognition, target following and the process such as screening, target alert, is effectively increased in actual environment the recognition efficiency to mobile target and enhances under complex environment detection and the trace ability to moving target；
Second: in application, make driver utilize monocular cam that driver can be assisted to be identified spacing and measure when driving, effectively prompting driver maintains safe distance, and reduces rear-end collision and occurs；
The second, this algorithm is simple and practical, both can download as independent cell phone software and use, drive recorder can also be encapsulated into, allow drive recorder have the function of anti-collision warning, really become the helper of driver, have only to Android phone or PC platform and can realize, plus common camera, the distance measurement function that tests the speed, without installing other hardware device additional, in today that smart mobile phone and vehicle-mounted Android navigation system are in full flourish, with low cost, popularization is good, and visualization is good.
Accompanying drawing explanation
Fig. 1 is the method flow schematic diagram of the present invention；
Fig. 2 is the workflow schematic diagram during the present invention applies.
Detailed description of the invention
Below in conjunction with the accompanying drawings and embodiment being described in further detail based on machine vision and the vehicle detection of cascade classifier and method for early warning the present invention.
As shown in Figure 1 and Figure 2, the present invention is to provide a kind of intelligent distance-measuring method based on monocular cam, catching front vehicles by machine vision and carry out image procossing, obtaining the change in location information of automobile thus reach the function of pre-vehicle rear-end collision prevention, the concrete grammar of the present invention may be summarized to be:
(1) positive and negative sample training cascade classifier is utilized；
(2) real time imaging obtains；
(3) Image semantic classification；
(4) target recognition；
(5) target following and screening；
(6) target alert.
Wherein the acquisition of image refers to by photographic head getting frame picture, call OPENCV storehouse and process picture, concrete handling process is as follows: first being processed by the picture scaling obtained, scale value needs to consider accuracy of identification and recognition rate here, when scale value increases, accuracy of identification reduces, recognition rate is accelerated, and when scale value reduces, accuracy of identification increases, recognition rate reduces, it is possible to by checking that recognition effect determines scale value.Need afterwards the picture after scaling is carried out gray proces, be used for improving recognition rate.Owing to the picture contrast of gray proces is the highest, it is unfavorable for the identification of grader, so needing gray scale picture is carried out gray-level histogram equalization before grader identification, in debugging finds picture, some is not required to identify, the left and right frame of such as picture and upper ledge, because the vehicle that these regions do not have vehicle appearance or appearance does not results in threat, reduce area-of-interest to accelerate recognition speed so needing to tint this region before grader identification.The picture now processed may be used for grader identification, passes through the training to Various Classifiers on Regional and detection, and after considering accuracy of identification and recognition rate, we determine to use LBP grader to carry out detection identification to processing picture.
The algorithm of target recognition specifically includes that goal ordering algorithm, vehicle front detection algorithm, the tracing algorithm of vehicle, the smoothing processing algorithm of identification width of frame.Wherein goal ordering refers to be ranked up to the distance of oneself according to target, the distance of target is corresponding with the y-coordinate value identifying vehicle in machine vision, distance is the most remote, and it identifies that the y value of frame is the biggest, it is possible to identified that by contrast the y-coordinate value of frame is ranked up；Vehicle front detection refers to the vehicle that oneself may be threatened by front in longitudinal driving process, here by point at the bottom of as triangle two of the transverse width in oneself vehicle width correspondence video known, road infinite point in video is considered as vertex of a triangle, connect three points triangle, when identifying that frame is considered as front vehicles when colliding with this detection triangle, the front vehicle to oneself being likely to result in threat is filtered out with this, simultaneously according to identifying the sequence of frame y-coordinate value, find apart from oneself nearest vehicle；Car tracing algorithm refers to be contrasted by the identification frame coordinate of previous frame with next frame, uses Euclidean distance to calculate the distance of two coordinates, when distance is considered as same car in threshold range, and take that threshold value is vehicle identification width of frame here 1/2nd；Owing to identification process is constantly in dynamically change, identify that width can not the most just comprise vehicle, so needing, to identifying that width is smoothed, to use moving average filtering processing method here, to the vehicle tracked by the identification width calculation meansigma methods of its front four frames.Existing in order to prevent the identification that some identification errors are bigger from outlining, need to carry out pretreatment before to identification frame smoothing processing, can use the identification width of frame contrast identifying width of frame and previous frame of present frame, when two frame identification width of frame differences exceed the 0.1 of previous frame width, this frame identification width of frame is adjusted to the meansigma methods of present frame identification width of frame and previous frame identification width of frame.
Identify distance aspect, we have employed the mode of high-definition camera subsidiary convex lens amplification and obtain picture and in terms of picture pretreatment, to increase, the mode of distant place regional area amplification identified distance, the identification of vehicle in 80 meters has been met by program debugging, owing to vehicle influential on us in actual life is not over 80 meters, solve the problem identifying that distance is not enough.
Early warning system have employed three grades of early warning mechanisms: trigger three grades of early warning when front detects vehicle；Two grades of early warning are triggered when front is less than 20 meters away from oneself nearest vehicle distances；When front vehicles relative to the speed of oneself it may happen that collision time less than 2 seconds time trigger one-level early warning.Wherein the calculation of distance is: use the photographic head of fixed resolution to simulate the relational expression identifying width of frame with actual range through repeatedly measured data, here the relational expression that we obtain:, wherein dist represents actual range, width is the width identifying frame, k is a constant coefficient relevant to resolution ratio of camera head, and the identification width of frame after bringing smoothing processing again into by the relational expression obtained can obtain the actual range identifying vehicle.The speed calculation mode of identification vehicle is: the vehicle distances that present frame is followed the tracks of deducts the distance of vehicle of the same tracking of previous frame as displacement difference, the system time of present frame is deducted the system time of previous frame as time difference, the instantaneous velocity of vehicle can be obtained by displacement difference divided by time difference.The Collision time calculation following the tracks of vehicle possible with front is: first determine whether currently to follow the tracks of the speed of vehicle, if negative, represent and follow the tracks of vehicle away from oneself, without calculating possible collision time, if just, represent that tracking vehicle, near oneself, can use present frame to follow the tracks of the distance of vehicle and can obtain the time of possible collision divided by its instantaneous velocity.
The present invention has only to Android phone or PC platform can realize, plus common camera, the distance measurement function that tests the speed, it is not necessary to install other hardware device additional.In today that smart mobile phone and vehicle-mounted Android navigation system are in full flourish, with low cost, popularization is good, and visualization is good.
Embodiment 1, specific implementation process:
With reference to Fig. 1, the method utilizing the present invention carries out the job step of intelligent distance-measuring early warning in the process of moving and is performed as follows: step one: the auto graph of the most known vehicle through editing carries out the training of Adaboost, obtain the characteristic information of automobile, and preserve into xml document；The Principle of Process of training is as follows:
A: by the study of N number of training sample being obtained first Weak Classifier；
B: the sample of misclassification constitutes together with other new data a new N number of training sample, by obtaining second Weak Classifier to the study of this sample；
C.: the sample of 1 and 2 all misclassifications is constituted another new N number of training sample, by the study of this sample obtains the 3rd Weak Classifier plus other new samples；
D: training sample always, until the optimal Weak Classifier selected meets predetermined condition, terminates training process.
Training result is preserved into XML extensible markup language, makes it have structural for labelling e-file, it is provided that unified method describes and exchange the structural data independent of application program or supplier.
Step 2: image acquisition and pretreatment, starts monocular cam and gathers image information, carries out denoising, smothing filtering pretreatment to gathering image information；
Step 3: utilize the XML file trained in step one, carries out vehicle identification to the image obtained, is positioned by each vehicle identified；
Step 4: determine the vehicle of dead ahead, carries out pixel analysis measurement to image, calculates the vehicle relative distance information to this car, and processing speed is preferably 20 frames/second；
Step 5: contrast with the range information of former frame, obtains relative velocity；
Step 6: utilize the relative distance of current time divided by relative velocity, calculate the anticipated collision time, if the value obtained is less than setting threshold value, then triggers anti-collision warning alarm or output signal to brakes and carry out slowing down or braking, if above threshold value, then continue detection.
The present invention has only to Android phone or PC platform can realize, plus common camera, the distance measurement function that tests the speed, it is not necessary to install other hardware device additional.In today that smart mobile phone and vehicle-mounted Android navigation system are in full flourish, with low cost, popularization is good, and visualization is good.
As described above, it is only presently preferred embodiments of the present invention, can not limit, with this, the scope that the present invention implements, the simple equivalence i.e. in every case made according to scope of the present invention patent and invention description content changes and modifies, the most still remaining within the scope of the patent.
Claims (7)
1. based on machine vision and the vehicle detection of cascade classifier and method for early warning, it is characterised in that: comprise the steps:
Utilize the cascade classifier training carrying out positive negative sample through the auto graph of the most known vehicle of editing；
The mathematical model identifying width of frame with reality distance is set up by a large amount of measured datas；
Real time imaging obtains；
Image semantic classification, confines area-of-interest；
Area-of-interest is carried out target recognition；
Recognition result is carried out data analysis, thus screens and follow the tracks of target；
It is identified the process of frame moving average filtering to following the tracks of target；
Calculate the actual range identifying target, calculated the translational speed of vehicle by frame difference time and actual range difference, obtain with this it may happen that time of colliding show early warning information.
The most according to claim 1 based on machine vision and the vehicle detection of cascade classifier with method for early warning, it is characterised in that described grader LBP based on vehicle tail features training obtains.
The most according to claim 1 based on machine vision and the vehicle detection of cascade classifier with method for early warning, it is characterized in that in step B, mathematical model is set up in the following manner: by a large amount of measured datas, obtain in entity and image width and the dependency relation of distance and carry out fitting function error analysis, determining range finding mathematical model.
The most according to claim 1 based on machine vision and the vehicle detection of cascade classifier with method for early warning, it is characterised in that in step D, to carry out by following procedure:
Original image is converted into OPENCV accessible Mat object；
To mat object scaling and gray proces；
Gray level image is carried out gray-level histogram equalization；
Use mask shade image that regions of non-interest realizes image to cover, thus reduce area-of-interest to reduce the complex environment interference to grader identification.
It is the most according to claim 1 based on machine vision and the vehicle detection of cascade classifier with method for early warning, it is characterised in that: step E includes that goal ordering and vehicle front detect process, specific as follows:
Goal ordering refers to be ranked up to the distance of oneself according to target, and the distance of target is corresponding with the y-coordinate value identifying vehicle in machine vision, and distance is the most remote, and it identifies that the y value of frame is the biggest, is identified that by contrast the y-coordinate value of frame is ranked up；
Vehicle front detection refers to the vehicle that oneself may be threatened by front in longitudinal driving process, here by point at the bottom of as triangle two of the transverse width in oneself vehicle width correspondence video known, road infinite point in video is considered as vertex of a triangle, connect three points triangle, when identifying that frame is considered as front vehicles when colliding with this detection triangle, the front vehicle to oneself being likely to result in threat is filtered out with this, simultaneously according to identifying the sequence of frame y-coordinate value, find apart from oneself nearest vehicle.
The most according to claim 1 based on machine vision and the vehicle detection of cascade classifier with method for early warning, it is characterized in that: in step F, contrasted by the identification frame coordinate of previous frame with next frame, Euclidean distance is used to calculate the distance of two coordinates, when distance is considered as same car in threshold range, take that threshold value is vehicle identification width of frame here 1/2nd；Use moving average filtering processing method to identifying that width is smoothed, to the vehicle tracked by the identification width calculation meansigma methods of its front four frames.
The most according to claim 6 based on machine vision and the vehicle detection of cascade classifier with method for early warning, it is characterized in that: before to identification frame smoothing processing, carry out pretreatment, detailed process is as follows: by the identification width of frame contrast identifying width of frame and previous frame of present frame, when two frame identification width of frame differences exceed the 0.1 of previous frame width, this frame identification width of frame is adjusted to the meansigma methods of present frame identification width of frame and previous frame identification width of frame.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510236672.7A CN106295459A (en) | 2015-05-11 | 2015-05-11 | Based on machine vision and the vehicle detection of cascade classifier and method for early warning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510236672.7A CN106295459A (en) | 2015-05-11 | 2015-05-11 | Based on machine vision and the vehicle detection of cascade classifier and method for early warning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106295459A true CN106295459A (en) | 2017-01-04 |
Family
ID=57630911
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510236672.7A Pending CN106295459A (en) | 2015-05-11 | 2015-05-11 | Based on machine vision and the vehicle detection of cascade classifier and method for early warning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106295459A (en) |
Cited By (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106952308A (en) * | 2017-04-01 | 2017-07-14 | 上海蔚来汽车有限公司 | The location determining method and system of moving object |
| CN107563282A (en) * | 2017-07-25 | 2018-01-09 | 大圣科技股份有限公司 | For unpiloted recognition methods, electronic equipment, storage medium and system |
| CN107862287A (en) * | 2017-11-08 | 2018-03-30 | 吉林大学 | A kind of front zonule object identification and vehicle early warning method |
| CN107886043A (en) * | 2017-07-20 | 2018-04-06 | 吉林大学 | The vehicle front-viewing vehicle and pedestrian anti-collision early warning system and method for visually-perceptible |
| CN108417089A (en) * | 2018-03-14 | 2018-08-17 | 杭州分数科技有限公司 | Traffic safety method for early warning, apparatus and system |
| CN109664820A (en) * | 2018-09-25 | 2019-04-23 | 平安科技（深圳）有限公司 | Driving reminding method, device, equipment and storage medium based on automobile data recorder |
| CN109817009A (en) * | 2018-12-31 | 2019-05-28 | 天合光能股份有限公司 | A method for obtaining dynamic traffic information required for unmanned driving |
| CN109859489A (en) * | 2019-01-30 | 2019-06-07 | 驭势科技（北京）有限公司 | A kind of spacing estimation method, device, mobile unit and storage medium |
| CN110222652A (en) * | 2019-06-10 | 2019-09-10 | 湖南国科微电子股份有限公司 | Pedestrian detection method, device and electronic equipment |
| CN110356392A (en) * | 2019-05-31 | 2019-10-22 | 惠州市德赛西威汽车电子股份有限公司 | A kind of intelligent householder method of safety traffic and system |
| CN110555347A (en) * | 2018-06-01 | 2019-12-10 | 杭州海康威视数字技术股份有限公司 | Method, device and electronic equipment for vehicle target recognition with dangerous cargo behavior |
| CN110633637A (en) * | 2019-08-09 | 2019-12-31 | 河海大学常州校区 | Assisted driving method based on Haar-Like feature extraction algorithm and gray value difference analysis |
| CN111448597A (en) * | 2017-10-10 | 2020-07-24 | 日产自动车株式会社 | Driving control method and driving control device |
| CN111508274A (en) * | 2020-03-17 | 2020-08-07 | 东软集团股份有限公司 | Collision early warning method, device and equipment |
| CN111753626A (en) * | 2019-03-28 | 2020-10-09 | 通用汽车环球科技运作有限责任公司 | Attention area identification for enhanced sensor-based detection in a vehicle |
| CN111914627A (en) * | 2020-06-18 | 2020-11-10 | 广州杰赛科技股份有限公司 | A vehicle identification and tracking method and device |
| CN112349142A (en) * | 2020-10-29 | 2021-02-09 | 深圳市元征科技股份有限公司 | Vehicle rear-end collision early warning method and related equipment thereof |
| CN111422190B (en) * | 2020-04-03 | 2021-08-31 | 北京四维智联科技有限公司 | Forward collision early warning method and system for rear car loader |
| WO2021217372A1 (en) * | 2020-04-27 | 2021-11-04 | 深圳市大疆创新科技有限公司 | Control method and device for movable platform |
| CN113947946A (en) * | 2021-10-15 | 2022-01-18 | 华东交通大学 | Port area traffic safety monitoring method based on Internet of vehicles V2X and video fusion |
| CN114092762A (en) * | 2021-11-10 | 2022-02-25 | 中国科学院空天信息创新研究院 | Method, device and electronic device for generating training samples for remote sensing image classification |
| CN120063208A (en) * | 2025-04-28 | 2025-05-30 | 中国科学院长春光学精密机械与物理研究所 | Vehicle-mounted theodolite-based intelligent tracking system and method for spacecraft parachute |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101574970A (en) * | 2009-03-06 | 2009-11-11 | 北京中星微电子有限公司 | Method and device for monitoring vehicle to change lane |
| US20100061624A1 (en) * | 2008-09-11 | 2010-03-11 | Wesley Kenneth Cobb | Detecting anomalous events using a long-term memory in a video analysis system |
| CN102642510A (en) * | 2011-02-17 | 2012-08-22 | 汽车零部件研究及发展中心有限公司 | Image-based vehicle anti-collision early warning method |
-
2015
- 2015-05-11 CN CN201510236672.7A patent/CN106295459A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100061624A1 (en) * | 2008-09-11 | 2010-03-11 | Wesley Kenneth Cobb | Detecting anomalous events using a long-term memory in a video analysis system |
| CN101574970A (en) * | 2009-03-06 | 2009-11-11 | 北京中星微电子有限公司 | Method and device for monitoring vehicle to change lane |
| CN102642510A (en) * | 2011-02-17 | 2012-08-22 | 汽车零部件研究及发展中心有限公司 | Image-based vehicle anti-collision early warning method |
Non-Patent Citations (1)
| Title |
|---|
| 周天弋: "基于计算机视觉的车辆与前方车距检测", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (33)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106952308B (en) * | 2017-04-01 | 2020-02-28 | 上海蔚来汽车有限公司 | Method and system for determining position of moving object |
| CN106952308A (en) * | 2017-04-01 | 2017-07-14 | 上海蔚来汽车有限公司 | The location determining method and system of moving object |
| WO2018177159A1 (en) * | 2017-04-01 | 2018-10-04 | 上海蔚来汽车有限公司 | Method and system for determining position of moving object |
| CN107886043B (en) * | 2017-07-20 | 2022-04-01 | 吉林大学 | Vision-aware anti-collision early warning system and method for forward-looking vehicles and pedestrians of automobile |
| CN107886043A (en) * | 2017-07-20 | 2018-04-06 | 吉林大学 | The vehicle front-viewing vehicle and pedestrian anti-collision early warning system and method for visually-perceptible |
| CN107563282A (en) * | 2017-07-25 | 2018-01-09 | 大圣科技股份有限公司 | For unpiloted recognition methods, electronic equipment, storage medium and system |
| CN111448597B (en) * | 2017-10-10 | 2022-06-10 | 日产自动车株式会社 | Driving control method and driving control device |
| CN111448597A (en) * | 2017-10-10 | 2020-07-24 | 日产自动车株式会社 | Driving control method and driving control device |
| CN107862287A (en) * | 2017-11-08 | 2018-03-30 | 吉林大学 | A kind of front zonule object identification and vehicle early warning method |
| CN108417089A (en) * | 2018-03-14 | 2018-08-17 | 杭州分数科技有限公司 | Traffic safety method for early warning, apparatus and system |
| CN110555347A (en) * | 2018-06-01 | 2019-12-10 | 杭州海康威视数字技术股份有限公司 | Method, device and electronic equipment for vehicle target recognition with dangerous cargo behavior |
| CN109664820A (en) * | 2018-09-25 | 2019-04-23 | 平安科技（深圳）有限公司 | Driving reminding method, device, equipment and storage medium based on automobile data recorder |
| CN109817009A (en) * | 2018-12-31 | 2019-05-28 | 天合光能股份有限公司 | A method for obtaining dynamic traffic information required for unmanned driving |
| CN109859489A (en) * | 2019-01-30 | 2019-06-07 | 驭势科技（北京）有限公司 | A kind of spacing estimation method, device, mobile unit and storage medium |
| CN111753626A (en) * | 2019-03-28 | 2020-10-09 | 通用汽车环球科技运作有限责任公司 | Attention area identification for enhanced sensor-based detection in a vehicle |
| CN111753626B (en) * | 2019-03-28 | 2023-09-19 | 通用汽车环球科技运作有限责任公司 | Attention area identification for enhanced sensor-based detection in a vehicle |
| CN110356392A (en) * | 2019-05-31 | 2019-10-22 | 惠州市德赛西威汽车电子股份有限公司 | A kind of intelligent householder method of safety traffic and system |
| CN110222652B (en) * | 2019-06-10 | 2021-07-27 | 湖南国科微电子股份有限公司 | Pedestrian detection method and device and electronic equipment |
| CN110222652A (en) * | 2019-06-10 | 2019-09-10 | 湖南国科微电子股份有限公司 | Pedestrian detection method, device and electronic equipment |
| CN110633637B (en) * | 2019-08-09 | 2023-05-16 | 河海大学常州校区 | Assisted driving method based on Haar-Like feature extraction algorithm and gray value difference analysis |
| CN110633637A (en) * | 2019-08-09 | 2019-12-31 | 河海大学常州校区 | Assisted driving method based on Haar-Like feature extraction algorithm and gray value difference analysis |
| CN111508274A (en) * | 2020-03-17 | 2020-08-07 | 东软集团股份有限公司 | Collision early warning method, device and equipment |
| CN111422190B (en) * | 2020-04-03 | 2021-08-31 | 北京四维智联科技有限公司 | Forward collision early warning method and system for rear car loader |
| CN114096931B (en) * | 2020-04-27 | 2025-02-18 | 深圳市大疆创新科技有限公司 | Control method and device for movable platform |
| WO2021217372A1 (en) * | 2020-04-27 | 2021-11-04 | 深圳市大疆创新科技有限公司 | Control method and device for movable platform |
| CN114096931A (en) * | 2020-04-27 | 2022-02-25 | 深圳市大疆创新科技有限公司 | Control method and device for movable platform |
| CN111914627A (en) * | 2020-06-18 | 2020-11-10 | 广州杰赛科技股份有限公司 | A vehicle identification and tracking method and device |
| CN112349142A (en) * | 2020-10-29 | 2021-02-09 | 深圳市元征科技股份有限公司 | Vehicle rear-end collision early warning method and related equipment thereof |
| CN113947946A (en) * | 2021-10-15 | 2022-01-18 | 华东交通大学 | Port area traffic safety monitoring method based on Internet of vehicles V2X and video fusion |
| CN114092762A (en) * | 2021-11-10 | 2022-02-25 | 中国科学院空天信息创新研究院 | Method, device and electronic device for generating training samples for remote sensing image classification |
| CN114092762B (en) * | 2021-11-10 | 2024-11-26 | 中国科学院空天信息创新研究院 | Remote sensing image classification training sample generation method, device and electronic equipment |
| CN120063208A (en) * | 2025-04-28 | 2025-05-30 | 中国科学院长春光学精密机械与物理研究所 | Vehicle-mounted theodolite-based intelligent tracking system and method for spacecraft parachute |
| CN120063208B (en) * | 2025-04-28 | 2025-08-01 | 中国科学院长春光学精密机械与物理研究所 | Vehicle-mounted theodolite-based intelligent tracking system and method for spacecraft parachute |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106295459A (en) | Based on machine vision and the vehicle detection of cascade classifier and method for early warning | |
| CN112417967B (en) | Obstacle detection method, obstacle detection device, computer device, and storage medium | |
| CN110163904B (en) | Object labeling method, movement control method, device, equipment and storage medium | |
| EP3229041B1 (en) | Object detection using radar and vision defined image detection zone | |
| CN114495064B (en) | Vehicle surrounding obstacle early warning method based on monocular depth estimation | |
| US7327855B1 (en) | Vision-based highway overhead structure detection system | |
| CN109085829B (en) | Dynamic and static target identification method | |
| CN109087510A (en) | traffic monitoring method and device | |
| US10769420B2 (en) | Detection device, detection method, computer program product, and information processing system | |
| CN107463890A (en) | A kind of Foregut fermenters and tracking based on monocular forward sight camera | |
| CN114155511A (en) | Environmental information acquisition method for automatically driving automobile on public road | |
| CN114296095A (en) | Method, device, vehicle and medium for extracting effective target of automatic driving vehicle | |
| KR20230174730A (en) | Method for determining an approximate object position of a dynamic object, computer program, device and vehicle | |
| CN114084129A (en) | Fusion-based vehicle automatic driving control method and system | |
| Berriel et al. | A particle filter-based lane marker tracking approach using a cubic spline model | |
| CN114120255A (en) | Target identification method and device based on laser radar speed measurement | |
| CN118928462B (en) | Multi-dimensional perception automatic driving avoidance method and device | |
| CN116246143A (en) | A Fusion Perception Method Based on Millimeter Wave Radar and Monocular Camera | |
| CN117173666A (en) | An automatic driving target recognition method and system for unstructured roads | |
| CN106503698A (en) | A kind of signalized intersections queuing vehicle static state spacing method for rapidly estimating | |
| CN105730330B (en) | Driving safety system and barrier screening method thereof | |
| US20250076880A1 (en) | High-definition mapping | |
| CN111881748B (en) | A lane line visual recognition method and system based on VBAI platform modeling | |
| CN119479361A (en) | Vehicle detection and collision warning method, device, system, and storage medium | |
| CN116587978A (en) | A collision warning method and system based on a vehicle-mounted display screen |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170104 |