CN114758511A - Sports car overspeed detection system, method, electronic equipment and storage medium - Google Patents
Sports car overspeed detection system, method, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN114758511A CN114758511A CN202210666406.8A CN202210666406A CN114758511A CN 114758511 A CN114758511 A CN 114758511A CN 202210666406 A CN202210666406 A CN 202210666406A CN 114758511 A CN114758511 A CN 114758511A
- Authority
- CN
- China
- Prior art keywords
- sports car
- picture
- splicing
- overspeed detection
- sports
- 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
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/052—Detecting movement of traffic to be counted or controlled with provision for determining speed or overspeed
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/017—Detecting movement of traffic to be counted or controlled identifying vehicles
- G08G1/0175—Detecting movement of traffic to be counted or controlled identifying vehicles by photographing vehicles, e.g. when violating traffic rules
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/052—Detecting movement of traffic to be counted or controlled with provision for determining speed or overspeed
- G08G1/054—Detecting movement of traffic to be counted or controlled with provision for determining speed or overspeed photographing overspeeding vehicles
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention provides a sports car overspeed detection system, a sports car overspeed detection method, electronic equipment and a storage medium, and belongs to the technical field of deep learning. The sports car overspeed detection system comprises a video acquisition module, a sports car identification module and a sports car overspeed detection module; the video acquisition module, the sports car identification module and the sports car overspeed detection module are sequentially connected; the video acquisition module is used for acquiring video stream information, decoding and preprocessing the video stream information; the sports car identification module is used for constructing a sports car identification model and identifying sports cars; the sports car overspeed detection module is used for realizing the overspeed detection of the sports car. The system also comprises a data processing and transmitting module which is used for transmitting the racing car overspeed detection result to an electronic police system, and the electronic police system is linked to upload the violation video and the violation racing car license plate record to a traffic police platform to be used as a punishment basis. The invention solves the problems that the traditional manual detection method for the vehicle running violation wastes time and labor, the automatic detection method has more investment, great damage to road surfaces and the like, and ensures the normal order of urban traffic.
Description
Technical Field
The application relates to a detection method, in particular to a sports car overspeed detection system, a sports car overspeed detection method, electronic equipment and a storage medium, and belongs to the technical field of deep learning.
Background
Super sports car often possesses the engine of big discharge capacity and a large amount of turbines, combines high strength and lightweight combined material, can realize comparing in daily car's faster moving speed, if drive sports car with the speed extremely fast in urban area street, along with unique engine sound, this will bring very big hidden danger for the traffic safety in city, untimely give certain punishment to this type of driver, probably causes serious traffic accident.
Because most of traffic enforcement equipment installed in the current urban road is electric police checkpoints, vehicles which run red lights, rule-breaking pressing lines and are not tied with safety belts at the intersection can only be supervised and punished, overspeed roadsters running on the road can not be detected and disposed, and common traffic police field supervision methods can not be applied to urban roads in a large range. The electric police with the speed measuring function is required to be provided with external field speed measuring equipment such as a coil and a radar, the coil needs to be buried under the ground, the maintenance is difficult, the construction is forbidden on the road surfaces such as a bridge overpass, and the like, the manufacturing cost of the radar is high, and large capital investment is required.
The traditional manual punishment work of the overspeed behavior of the sports car needs to arrange a traffic police to carry out field supervision on intersections and road sections of urban roads and record violation videos. Traditional sports car overspeed behavior automated inspection punishment need combine front end coil or radar equipment to go on, punish violating regulations through the electronic police system. The method has the defects of limited number of traffic police personnel, small jurisdiction area, high labor cost and the condition of manual misjudgment. The coil needs to be buried under the ground, so that the maintenance is difficult, the construction is forbidden on the road surfaces such as bridge overpass, and the like, and the radar has high cost and needs to invest larger funds.
Aiming at the situation, if the camera and the overspeed detection algorithm mounted on the roadside can be relied on, the overspeed monitoring algorithm of pure vision is adopted, the damage to the road surface and the fund investment can be reduced, the violation video evidence is kept and the electric police equipment is connected in parallel for snapshot when the sports car runs at an overspeed, and the illegal sports car can be effectively punished.
Therefore, a racing car overspeed detection method is needed to solve the problems that the traditional manual racing car violation detection method wastes time and labor, the automatic detection method has more investment, the road surface is greatly damaged and the like, and ensure the normal order of urban traffic.
Disclosure of Invention
The following presents a simplified summary of the invention in order to provide a basic understanding of some aspects of the invention. It should be understood that this summary is not an exhaustive overview of the invention. It is not intended to determine the key or critical elements of the present invention, nor is it intended to limit the scope of the present invention. Its sole purpose is to present some concepts in a simplified form as a prelude to the more detailed description that is discussed later.
In view of the above, the invention provides a sports car overspeed detection system, a method, an electronic device and a storage medium, which are used for solving the technical problems that the traditional sports car violation manual detection method in the prior art is time-consuming and labor-consuming, the automatic detection method is high in investment and great in damage to road surfaces.
The first scheme is as follows: a sports car overspeed detection system comprises a video acquisition module, a sports car identification module and a sports car overspeed detection module; the video acquisition module, the sports car identification module and the sports car overspeed detection module are sequentially connected;
the video acquisition module is used for acquiring video stream information, decoding and preprocessing the video stream information;
the sports car identification module is used for constructing a sports car identification model and identifying sports cars;
The sports car overspeed detection module is used for realizing the overspeed detection of the sports car.
Preferably, the system also comprises a data processing and transmitting module, wherein the input end of the data processing and transmitting module is connected with the output end of the sports car overspeed detection module, and the output end of the data processing and transmitting module is connected with an electronic police system; the system is used for transmitting the racing car overspeed detection result to an electronic police system, and the electronic police system is linked to upload the violation video and the license plate record of the violation racing car to a traffic police platform to serve as a punishment basis.
Scheme II: a sports car overspeed detection method comprises the following steps:
s1, acquiring traffic video stream information and preprocessing the video;
s2, constructing a sports car identification model;
s3, judging whether the sports car is overspeed, the specific method is as follows: the method comprises the following steps:
s31, drawing virtual detection lines perpendicular to the vehicle type direction by drawing two virtual detection lines perpendicular to the road lines on the road line and the lane dividing line; assume detection line 1: y1= Ax1+ B, detection line 2: y2= Cx2+ D, and passes through the detection line 1 and then 2 according to the normal form path of the vehicle;
s32, acquiring coordinates of a vehicle surrounding rectangular frame; since the coordinates output by YOLOv3 are the center point coordinates of the bounding car frame and the width and length of the rectangular frame, assuming that the parameters of the rectangular frame b which is called to currently bound the car are (b.x, b.y, b.w, b.h), the coordinates (upper left, upper right, lower left, lower right) of the four corners of the current rectangular frame are (b.x-b.w/2, b.y-b.h/2), (b.x + b.w/2, b.y-b.h/2), (b.x-b.w/2, b.y + b.h/2), (b.x + b.w/2, b.y + b.h/2);
S33, detecting the overspeed of the sports car.
Preferably, the method for acquiring traffic video stream information and preprocessing the video includes performing color space conversion and image filtering and denoising on a single frame of picture.
Preferably, the method for constructing the sports car recognition model comprises two parts of model construction and model training,
model construction: changing Darknet-53 model architecture codes in Visual Studio, and completing successful compiling of the model;
model training: and constructing a sports car data set, inputting the constructed sports car data set into an improved YOLOv3 model, inputting pixel data of an input picture into the model, comparing output with expected output, calculating an error by using a cost function, iteratively updating the weight of the model, and reducing the training loss rate of the model to the minimum value so as to finish the final weight determination of the model.
Preferably, the method for constructing the sports car data set comprises the following steps:
s21, acquiring 1000 sports car photos in different shooting scenes;
s22, splicing and expanding sports car pictures, wherein the splicing and expanding method comprises the steps of 4-picture splicing and expanding, 10-picture splicing and 16-picture splicing and expanding;
s22.4, splicing and expanding, comprising the following steps:
S221, processing the picture pixels into a pixel a and b in a unified mode;
s222, randomly extracting 4 pictures as a splicing material picture;
s223, 1/2 pictures are transversely spliced to obtain a picture a; transversely splicing the remaining 1/2 pictures to obtain a picture b;
s224, vertically splicing the picture a and the picture b to obtain a picture c, wherein the picture c is a training picture;
s23.10 the splicing and expanding method comprises the following steps:
s231, randomly extracting 10 pictures as a splicing material picture;
s232, splicing 8 pictures according to the splicing and expanding method of the 4 pictures to obtain a picture d and a picture e;
s233, adjusting the pixels of the picture d, the picture e and the rest two pictures to be a pixel b;
s234, splicing the rest 2 pictures with the picture d and the picture e according to the 4-picture splicing expansion method to obtain a picture f, wherein the picture f is a training picture;
s24.16 splicing and expanding, comprising the following steps:
s241, randomly extracting 16 pictures as a splicing material picture;
s242, splicing according to the splicing expansion method of the 4 pictures to obtain a picture h, a picture g, a picture i and a picture j;
and S234, splicing the picture h, the picture g, the picture i and the picture j according to the 4-picture splicing expansion method to obtain a picture k, wherein the picture k is a training picture.
Preferably, the method comprises the following steps: the method for detecting the overspeed of the sports car comprises the following steps:
S331, taking the coordinate of the upper right corner of the sports car as a reference, recording the current time t1 when the coordinate enters the virtual detection line 1 for the first time, namely y1 is more than or equal to A (b.x + b.w/2) + B and G is more than or equal to b.x + b.w/2 is more than or equal to H, and recording the coordinate of the upper right corner of the current sports car (the coordinate of the upper right corner of the current sports car is) (B) (2)
,
) (ii) a When the coordinate enters a virtual detection line 2 for the first time, namely y2 is not less than C (b.x + b.w/2) + D and G is not less than b.x + b.w/2 is not less than H, recording the current time t2 and the coordinate of the upper right corner of the current sports car (C) (b)
,
）；
S332, knowing coordinates according to an algorithm of moving actual distance of the sports car in the virtual detection line (
,
），（
,
) Corresponding actual coordinates can be obtained (
,
）、（
，
）；
S333, according to the actual coordinates (
,
）、（
，
) Calculating the moving actual distance F of the sports car in the virtual detection line,
；
and S334, calculating the average speed v = F/(t2-t1) of the sports car, and comparing the average speed of the sports car with the local speed limit to judge whether the current sports car is overspeed or not.
And a third scheme is as follows: an electronic device comprising a memory storing a computer program and a processor implementing the steps of a sports car overspeed detection method according to aspect one when executing the computer program.
The scheme four is as follows: a computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out a sports car overspeed detection method according to aspect one.
The invention has the following beneficial effects: the invention relies on the video shot by the camera mounted on the roadside, adopts the deep learning technology to identify and process the sports car, judges whether the sports car has overspeed behavior or not through the multidimensional image processing technology, if yes, stores the video evidence of the sports car violation, links with an electric police to take a candid photograph, and uploads the data to a traffic police platform. The method has the characteristics of simple deployment, higher precision, less investment, strong real-time performance and the like, can reduce the labor investment of field supervision traffic polices and the investment of coil radar purchasing and laying capital, and realizes fine management of cities. The invention solves the problems that the traditional manual detection method for the vehicle running violation wastes time and labor, the automatic detection method has more investment, great damage to road surfaces and the like, and ensures the normal order of urban traffic.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the application and together with the description serve to explain the application and not to limit the application. In the drawings:
FIG. 1 is a schematic diagram of the system architecture of the present invention;
FIG. 2 is a schematic flow chart of the method of the present invention;
FIG. 3 is a schematic diagram of the minimum bounding rectangle of the present invention;
FIG. 4 is a schematic diagram of drawing virtual detection lines according to the present invention;
FIG. 5 is a schematic diagram of the present invention for drawing virtual detection lines in conjunction with an actual road;
FIG. 6 is a schematic diagram of the present invention for determining racing car overspeed;
FIG. 7 is a diagram illustrating a fifth residual error network according to the present invention;
FIG. 8 is a schematic diagram of a network architecture of a vehicle identification model.
Detailed Description
In order to make the technical solutions and advantages of the embodiments of the present application more apparent, the following further detailed description of the exemplary embodiments of the present application with reference to the accompanying drawings makes it clear that the described embodiments are only a part of the embodiments of the present application, and are not exhaustive of all embodiments. It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict.
Embodiment 1, this embodiment is described with reference to fig. 1, and a sports car overspeed detection system includes a video acquisition module, a sports car identification module, and a sports car overspeed detection module; the video acquisition module, the sports car identification module and the sports car overspeed detection module are sequentially connected;
the video acquisition module is used for acquiring video stream information, decoding and preprocessing the video stream information;
the sports car identification module is used for constructing a sports car identification model and identifying sports cars;
The sports car overspeed detection module is used for realizing sports car overspeed detection.
The input end of the data processing and transmitting module is connected with the output end of the sports car overspeed detection module, and the output end of the data processing and transmitting module is connected with an electronic police system; the system is used for transmitting the sports car overspeed detection result to an electronic police system, and linking the electronic police system to upload the violation video and the violation sports car license plate record to a traffic police platform to serve as a punishment basis.
The processing flow among the modules in this embodiment is as follows: firstly, a video acquisition module acquires a real-time video of a road shot by a road side camera, then the real-time video is processed by a sports car identification module, a sports car frame in the video is selected by utilizing a trained sports car identification model, then the real-time video enters a sports car overspeed detection module, the average speed of the sports car passing through the road section is acquired, the average speed is compared with the local limited speed, when the fact that the sports car exceeds the highest speed limit is detected, video evidence violating regulations is recorded and is combined with electronic police equipment for snapshot, and the video evidence is uploaded to a traffic police platform through a data transmission module.
Example 2, this embodiment is described with reference to fig. 2 to 8, and a sports car overspeed detection method includes the steps of:
S1, traffic video stream information is obtained, and the video is preprocessed;
s11, accessing camera video stream information: and the RJ45 Ethernet cable is used for connecting the camera with the network interface of the edge computing gateway, and the edge computing gateway is arranged in the system software of the edge computing gateway and is accessed into the real-time video stream information acquired by the camera in a RTSP video stream address mode.
S12, video decoding: the method is characterized in that a file in an original video format is converted into a file in another video format through a compression technology. The most important coding and decoding standards in video streaming transmission are H.261, H.263 and H.264 of the International telecommunication Union.
S13, video preprocessing: and color space conversion and image filtering and denoising processing are carried out on the single-frame image, so that subsequent depth processing operation on the image is facilitated.
S2, constructing a sports car identification module; in this embodiment, an improved yollov 3 model is used to adjust the original structure of the Darknet-53 model, and referring to fig. 7, in the diagram part of the original network architecture, a convolutional layer of 1X1 and 512Filters is added to the fifth residual network structure, so as to improve the accuracy of identifying a sports car target far away from the camera.
The original Darknet-53 model structure is mainly composed of a series of convolution layers of 1x1 and 3x3 (each convolution layer is followed by a BN layer and a LeakyReLU), 53 Connected layers are in the network, so the model is called Darknet-53 (2 + 1x 2 + 1 + 2 x 2 + 1 + 8 + 2 + 1 + 4 x 2 + 1 = 53 according to the sequence number, and the final Connected layer is also calculated as the convolution layer, and the total number is 53).
The construction of the sports car recognition module comprises two parts of model construction and model training;
model construction: changing Darknet-53 model architecture codes in Visual Studio, and completing successful compiling of the model;
model training: and constructing a sports car data set, inputting the constructed sports car data set into an improved YOLOv3 model, inputting pixel data of an input picture into the model, comparing output with expected output, calculating an error by using a cost function, iteratively updating the weight of the model, and reducing the training loss rate of the model to the minimum value so as to finish the final weight determination of the model.
The method for constructing the sports car data set comprises the following steps:
s21, acquiring 1000 sports car photos in different shooting scenes;
s22, splicing and expanding sports car pictures, wherein the splicing and expanding method comprises the steps of 4-picture splicing and expanding, 10-picture splicing and 16-picture splicing and expanding;
s22.4, splicing and expanding, wherein the splicing and expanding method comprises the following steps:
s221, processing the picture pixels into a pixels and b pixels in a unified mode;
s222, randomly extracting 4 pictures as a splicing material picture;
s223, taking 1/2 pictures to transversely splice to obtain a picture a; transversely splicing the remaining 1/2 pictures to obtain a picture b;
s224, vertically splicing the picture a and the picture b to obtain a picture c, wherein the picture c is a training picture;
S23.10 splicing and expanding, comprising the following steps:
s231, randomly extracting 10 pictures as a splicing material picture;
s232, splicing 8 pictures according to the splicing and expanding method of the 4 pictures to obtain a picture d and a picture e;
s233, adjusting the pixels of the picture d, the picture e and the rest two pictures to be a pixel b;
s234, splicing the rest 2 pictures with the picture d and the picture e according to the 4-picture splicing expansion method to obtain a picture f, wherein the picture f is a training picture;
s24.16 splicing and expanding, comprising the following steps:
s241, randomly extracting 16 pictures as a splicing material picture;
s242, splicing according to the splicing expansion method of the 4 pictures to obtain a picture h, a picture g, a picture i and a picture j;
and S234, splicing the picture h, the picture g, the picture i and the picture j according to the 4-picture splicing expansion method to obtain a picture k, wherein the picture k is a training picture.
A neural network training process:
step 1, randomly initializing weights in neural network
And 2, sending the first group of input values to a neural network, so that the first group of input values are propagated through the network to obtain output values.
Step 3, comparing the output value with the expected output value, and calculating the error by using a cost function.
And 4, propagating the error back to the network, and setting the weight according to the information.
And 5, repeating the steps from 2 to 4 for each input value in the training set.
Step 6, when the whole training set is sent to the neural network, one epoch is completed, and then the epochs are repeated for a plurality of times); the model network comprises functions of feature extraction (feature extraction), candidate Region (Region deployment) extraction, border regression (bounding box regression), classification (classification) and the like, so that the comprehensive performance is greatly improved, and the detection speed is particularly obvious.
Referring to fig. 8, the overall vehicle identification model network infrastructure is illustrated; conv layers in the figure are improved YOLOv 3; region prolysals Networks in the graph for generating suggested regions (Region prolysals); the layer judges whether anchors belong to positive samples (positive) or negative samples (negative) through a softmax network, and corrects the anchors by utilizing bounding box regression (bounding box regression) to obtain accurate suggestions (positive); in the graph, Roi Pooling collects input feature maps (feature maps) and suggestions (suggestions), extracts the suggestion feature maps (suggestion features maps) after information is integrated, and sends the suggestion feature maps to a subsequent full-connection layer to judge the target category; in the Classification, the Classification of the suggestion (suggestion) is calculated by using the suggestion feature maps (suggestion feature maps), and meanwhile, the final accurate position of the detection frame is obtained by performing bounding box regression (bounding box regression) again.
The method comprises the steps of setting YOLOv3.cfg, obj.name and obj.data files of an improved YOLOv3 model by using Visual Studio of a configured environment, entering directory addresses of dark net.
S3, judging whether the sports car is overspeed, the concrete method is: the method comprises the following steps:
in order to realize the detection of the speed of the sports car, a virtual detection line is drawn by using OpenCV (open source vehicle controller) to provide a basis for a subsequent sports car overspeed detection algorithm, and referring to FIG. 4, two virtual detection lines perpendicular to a road line are drawn on the existing road line and a lane demarcation line. In the YOLOv3 model, the four output parameters of the rectangular frame are respectively the horizontal and vertical coordinates of the central point of the rectangular frame and the width and length of the rectangular frame, and the real-time numerical value of the sports car on the road section is deduced.
S31, drawing a virtual detection line perpendicular to the direction of the vehicle model, and assuming that the detection line 1: y1= Ax1+ B, detection line 2: y2= Cx2+ D, and passes through the detection line 1 and then 2 according to the normal form path of the vehicle;
s32, coordinates of a vehicle surrounding rectangular frame are obtained, and since the coordinates output by YOLOv3 are coordinates of a center point of the surrounding sports car frame and the width and the length of the rectangular frame, assuming that parameters of a rectangular frame b for calling the current surrounding sports car are (b.x, b.y, b.w., b.h), coordinates (upper left, upper right, lower left and lower right) of four corners of the current rectangular frame are (b.x-b.w/2, b.y-b.h/2), (b.x + b.w/2, b.y-b.h/2), (b.x-b.w/2, b.y + b.h/2), (b.x + b.w/2 and b.y + b.h/2).
S33, detecting racing car overspeed, taking the coordinate of the upper right corner of the racing car as a reference, when the coordinate enters a virtual line 1 for the first time, namely y1 is more than or equal to A (b.x + b.w/2) + B and G is more than or equal to b.x + b.w/2 and is less than or equal to H, recording the current time t1, and detecting the coordinate of the upper right corner of the current racing car (the coordinate of the upper right corner of the current racing car) (B + b.w/2)
,
) (ii) a When the coordinate enters the virtual line 2 for the first time, namely y2 ≦ C (b.x + b.w/2) + D and G ≦ b.x + b.w/2 ≦ H, the current time t2 is recorded, and the current upper right-hand coordinate of the sports car: (a) (b) is recorded
,
). According to the algorithm of the moving actual distance of the sports car in the virtual line, the known coordinates (
,
）、（
,
) Corresponding actual coordinates can be obtained (
,
）、（
，
) The actual moving distance F of the sports car in the virtual line can be calculated, so that the average speed v = F/(t2-t1) of the sports car can be calculated, and the average speed of the sports car is compared with the local speed limit to judge whether the current sports car is overspeed or not.
Specifically, the method comprises the following steps:
s331, taking the coordinate of the upper right corner of the sports car as a reference, recording the current time t1 when the coordinate enters the virtual detection line 1 for the first time, namely y1 is more than or equal to A (b.x + b.w/2) + B and G is more than or equal to b.x + b.w/2 is more than or equal to H, and recording the coordinate of the upper right corner of the current sports car (the coordinate of the upper right corner of the current sports car is) (B) (2)
,
) (ii) a When the coordinate enters a virtual detection line 2 for the first time, namely y2 is not less than C (b.x + b.w/2) + D and G is not less than b.x + b.w/2 is not less than H, recording the current time t2 and the coordinate of the upper right corner of the current sports car (C) (b) 
,
）；
S332. in the algorithm for moving the actual distance according to the sports car in the virtual detection line, the known coordinates (C:
,
），（
,
) Corresponding actual coordinates can be obtained (
,
）、（
，
）；
S333, calculating the actual moving distance F of the sports car in the virtual detection line according to the pixel coordinates, wherein the specific method is that the image coordinates are converted into world coordinates: suppose the pixel coordinate is (
，
) The camera coordinates are
，
，
) The actual coordinate is (
，
，
），
、
Is the front-end camera photosensitive device pixel size,
、
is the image midpoint coordinate.
Wherein the content of the first and second substances,
is the Z-coordinate value in the camera coordinates,
is a reference for the camera to be used,
for camera external reference, after the camera is erected and focused, the external reference can be obtained by a Zhang Zhengyou calibration method, H is the product of internal and external parameter matrixes of the camera, and H =is set
The formula for converting the image coordinates into world coordinates is abbreviated as
The values F3, G3 are obtained in the third row of the matrix,
is a rotation matrix
The inverse matrix of (c):
obtaining the values F3, G3 in the third row of the matrix can be derived
Further derive the result
If two point coordinates of the image are known (
,
），（
,
) In the speed measuring scene of a sports car
=0, the corresponding actual coordinates can be obtained: (
,
）、（
，
) Further calculate the direct distance between two actual coordinates
；
And S334, calculating the average speed v = F/(t2-t1) of the sports car, and comparing the average speed of the sports car with the local speed limit to judge whether the current sports car is overspeed or not.
Abbreviations and key term definitions of the invention:
conv: convolution, after all action points on the convolution kernel act on the original pixel points in sequence (multiplication), the output result is linearly superposed.
Bn (batchnorm): and the layer of the neural network forcibly pulls the input distribution which is gradually mapped to the nonlinear function and then is close to the extreme saturation region of the value-taking interval back to the standard normal distribution with the mean value of 0 and the variance of 1, so that the input value of the nonlinear transformation function falls into a region which is sensitive to input, and the problem of gradient disappearance is avoided.
LeakyReLU: unsaturated activation function, mathematical expression: y = max (0, x) + leak min (0, x).
Connected: neural network connectivity layer
Softmax: cross entropy, generally used in neural networks as the output layer of classification tasks.
Filters: and the filter can change the picture into multiple channels.
FPN: a feature pyramid model that combines multi-level features to solve multi-scale problems.
In embodiment 3, the computer device of the present invention may be a device including a processor, a memory, and the like, for example, a single chip microcomputer including a central processing unit, and the like. And the processor is used for implementing the steps of the recommendation method capable of modifying the relationship-driven recommendation data based on the CREO software when executing the computer program stored in the memory.
The Processor may be a Central Processing Unit (CPU), other general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic, discrete hardware components, etc. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The memory may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required by at least one function (such as a sound playing function, an image playing function, etc.), and the like; the storage data area may store data (such as audio data, a phonebook, etc.) created according to the use of the cellular phone, and the like. In addition, the memory may include high speed random access memory, and may also include non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), at least one magnetic disk storage device, a Flash memory device, or other volatile solid state storage device.
Example 4 computer-readable storage Medium example
The computer readable storage medium of the present invention may be any form of storage medium read by a processor of a computer device, including but not limited to non-volatile memory, ferroelectric memory, etc., on which a computer program is stored, which when read and executed by the processor of the computer device, may implement the steps of the above-described CREO software-based modeling method that can modify relationship-driven modeling data.
The computer program comprises computer program code which may be in the form of source code, object code, an executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U.S. disk, removable hard disk, magnetic diskette, optical disk, computer Memory, Read-Only Memory (ROM), Random Access Memory (RAM), electrical carrier wave signal, telecommunications signal, and software distribution medium, etc. It should be noted that the computer-readable medium may contain suitable additions or subtractions depending on the requirements of legislation and patent practice in jurisdictions, for example, in some jurisdictions, computer-readable media may not include electrical carrier signals or telecommunication signals in accordance with legislation and patent practice.
While the invention has been described with respect to a limited number of embodiments, those skilled in the art, having benefit of this description, will appreciate that other embodiments can be devised which do not depart from the scope of the invention as disclosed herein. Furthermore, it should be noted that the language used in the specification has been principally selected for readability and instructional purposes, and may not have been selected to delineate or circumscribe the inventive subject matter. Accordingly, many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the appended claims. The present invention has been disclosed in an illustrative rather than a restrictive sense, and the scope of the present invention is defined by the appended claims.
Claims (9)
1. A sports car overspeed detection system is characterized by comprising a video acquisition module, a sports car identification module and a sports car overspeed detection module; the video acquisition module, the sports car identification module and the sports car overspeed detection module are sequentially connected;
the video acquisition module is used for acquiring video stream information, decoding and preprocessing the video stream information;
the sports car identification module is used for constructing a sports car identification model and identifying a sports car;
The sports car overspeed detection module is used for realizing sports car overspeed detection.
2. The sports car overspeed detection system according to claim 1, further comprising a data processing transmission module, wherein an input end of said data processing transmission module is connected with an output end of said sports car overspeed detection module, and said output end is connected with an electronic police system; the electronic police system is linked with the electronic police system to upload violation videos and violation sports car license plate records to the traffic police platform.
3. A sports car overspeed detection method is characterized by comprising the following steps:
s1, traffic video stream information is obtained, and the video is preprocessed;
s2, constructing a sports car identification model;
s3, judging whether the sports car is overspeed, the specific method is as follows: the method comprises the following steps:
s31, drawing virtual detection lines perpendicular to the vehicle type direction by drawing two virtual detection lines perpendicular to the road lines on the road line and the lane dividing line; assume detection line 1: y1= Ax1+ B, detection line 2: y2= Cx2+ D, and passes through the detection line 1 and then 2 according to the normal form path of the vehicle;
s32, acquiring coordinates of a vehicle surrounding rectangular frame; since the coordinates output by YOLOv3 are the center point coordinates of the bounding car frame and the width and length of the rectangular frame, assuming that the parameters of the rectangular frame b which is called to currently bound the car are (b.x, b.y, b.w, b.h), the coordinates (upper left, upper right, lower left, lower right) of the four corners of the current rectangular frame are (b.x-b.w/2, b.y-b.h/2), (b.x + b.w/2, b.y-b.h/2), (b.x-b.w/2, b.y + b.h/2), (b.x + b.w/2, b.y + b.h/2);
S33, detecting the overspeed of the sports car.
4. The method as claimed in claim 3, wherein the method for obtaining traffic video stream information and preprocessing the video comprises performing color space conversion and image filtering denoising on a single frame picture.
5. A sports car overspeed detection method according to claim 4, characterized in that said method of building sports car identification model is, including model building and model training,
constructing a model: changing Darknet-53 model architecture codes in Visual Studio, and completing successful compiling of the model;
model training: and constructing a sports car data set, inputting the constructed sports car data set into an improved YOLOv3 model, inputting pixel data of an input picture into the model, comparing output with expected output, calculating an error by using a cost function, iteratively updating the weight of the model, and reducing the training loss rate of the model to the minimum value so as to finish the final weight determination of the model.
6. A sports car overspeed detection method according to claim 5, characterized in that said method of constructing sports car data sets is such as to comprise the following steps:
s21, acquiring 1000 sports car photos in different shooting scenes;
S22, splicing and expanding sports car pictures, wherein the concrete method comprises 4-picture splicing and expanding, 10-picture splicing and expanding and 16-picture splicing and expanding;
s22.4, splicing and expanding, comprising the following steps:
s221, processing the picture pixels into a pixel a and b in a unified mode;
s222, randomly extracting 4 pictures as a splicing material picture;
s223, 1/2 pictures are transversely spliced to obtain a picture a; transversely splicing the remaining 1/2 pictures to obtain a picture b;
s224, vertically splicing the picture a and the picture b to obtain a picture c, wherein the picture c is a training picture;
s23.10 the splicing and expanding method comprises the following steps:
s231, randomly extracting 10 pictures as a splicing material picture;
s232, splicing 8 pictures according to the splicing and expanding method of the 4 pictures to obtain a picture d and a picture e;
s233, adjusting the pixels of the picture d, the picture e and the rest two pictures to be a pixel b;
s234, splicing the rest 2 pictures with the picture d and the picture e according to the 4-picture splicing expansion method to obtain a picture f, wherein the picture f is a training picture;
s24.16 splicing and expanding, comprising the following steps:
s241, randomly extracting 16 pictures as a splicing material picture;
s242, splicing according to the splicing expansion method of the 4 pictures to obtain a picture h, a picture g, a picture i and a picture j;
And S234, splicing the picture h, the picture g, the picture i and the picture j according to a 4-picture splicing expansion method to obtain a picture k, wherein the picture k is a training picture.
7. A sports car overspeed detection method according to claim 6, characterized in that said sports car overspeed detection method is a method comprising the steps of:
s331, taking the coordinate of the upper right corner of the sports car as a reference, when the coordinate enters a virtual detection line 1 for the first time, namely y1 is less than or equal to A (b.x + b.w/2) + B and G is less than or equal to b.x + b.w/2 is less than or equal to H, recording the current time t1, and recording the coordinate of the upper right corner of the current sports car (a) (b.w/2) < H
,
) (ii) a When the coordinate enters the virtual detection line 2 for the first time, namely y2 ≦ C (b.x + b.w/2) + D and G ≦ b.x + b.w/2 ≦ H, recording the current time t2 and the current upper right-hand coordinate of the sports car: (a) (b) ((b) < 2 >)
,
）；
S332, according to the algorithm of the moving actual distance of the sports car in the virtual detection line, the known coordinates (C and D)
,
），（
,
) Corresponding actual coordinates can be obtained (
,
）、（
，
）；
S333, according to the actual coordinates (
,
）、（
，
) Calculating the moving actual distance F of the sports car in the virtual detection line,
；
and S334, calculating the average speed v = F/(t2-t1) of the sports car, and comparing the average speed of the sports car with the local speed limit to judge whether the current sports car is overspeed or not.
8. An electronic device, comprising a memory storing a computer program and a processor, wherein the processor implements the steps of a sports car overspeed detection method according to any of claims 3 to 7 when executing the computer program.
9. A computer-readable storage medium, on which a computer program is stored, which computer program, when being executed by a processor, is adapted to carry out a sports car overspeed detection method according to any one of claims 3 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210666406.8A CN114758511B (en) | 2022-06-14 | 2022-06-14 | Sports car overspeed detection system, method, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210666406.8A CN114758511B (en) | 2022-06-14 | 2022-06-14 | Sports car overspeed detection system, method, electronic equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114758511A true CN114758511A (en) | 2022-07-15 |
| CN114758511B CN114758511B (en) | 2022-11-25 |
Family
ID=82337226
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210666406.8A Active CN114758511B (en) | 2022-06-14 | 2022-06-14 | Sports car overspeed detection system, method, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114758511B (en) |
Citations (36)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105702048A (en) * | 2016-03-23 | 2016-06-22 | 武汉理工大学 | Automobile-data-recorder-based illegal lane occupation identification system and method for automobile on highway |
| CN106205134A (en) * | 2015-04-29 | 2016-12-07 | 上海宝康电子控制工程有限公司 | The comprehensive vehicle detecting system of highway |
| CN106530746A (en) * | 2016-10-28 | 2017-03-22 | 江苏信息职业技术学院 | Vehicle speed detection apparatus and detection method based on video technology |
| CN107958590A (en) * | 2017-11-20 | 2018-04-24 | 北京工业大学 | A kind of traffic offence crimping seizes lanes behavioral value acquisition system and method |
| US20180365858A1 (en) * | 2017-06-14 | 2018-12-20 | Hyundai Mobis Co., Ltd. | Calibration method and apparatus |
| CN109102702A (en) * | 2018-08-24 | 2018-12-28 | 南京理工大学 | Vehicle speed measuring method based on video encoder server and Radar Signal Fusion |
| CN109615870A (en) * | 2018-12-29 | 2019-04-12 | 南京慧尔视智能科技有限公司 | A kind of traffic detection system based on millimetre-wave radar and video |
| CN110178167A (en) * | 2018-06-27 | 2019-08-27 | 潍坊学院 | Crossing video frequency identifying method violating the regulations based on video camera collaboration relay |
| CN110399766A (en) * | 2019-01-28 | 2019-11-01 | 浙江浩腾电子科技股份有限公司 | Smoking testing and analysis system based on deep learning |
| JP2020042712A (en) * | 2018-09-13 | 2020-03-19 | 株式会社Ｊｖｃケンウッド | Video recording control device, video recording system, video recording method, and video recording control program |
| CN110909699A (en) * | 2019-11-28 | 2020-03-24 | 北京以萨技术股份有限公司 | Video vehicle non-guide driving detection method and device and readable storage medium |
| CN111009131A (en) * | 2019-12-05 | 2020-04-14 | 成都思晗科技股份有限公司 | High-order video intelligence parking system based on image recognition |
| CN111402593A (en) * | 2020-03-02 | 2020-07-10 | 哈尔滨工业大学 | Video traffic parameter acquisition method based on polynomial fitting |
| WO2020180620A1 (en) * | 2019-03-01 | 2020-09-10 | Amazon Technologies, Inc. | Identifying parking location using single camera reverse projection |
| CN111753797A (en) * | 2020-07-02 | 2020-10-09 | 浙江工业大学 | Vehicle speed measuring method based on video analysis |
| CN111915883A (en) * | 2020-06-17 | 2020-11-10 | 西安交通大学 | Road traffic condition detection method based on vehicle-mounted camera shooting |
| CN112037159A (en) * | 2020-07-29 | 2020-12-04 | 长安大学 | Cross-camera road space fusion and vehicle target detection tracking method and system |
| CN112102609A (en) * | 2020-03-10 | 2020-12-18 | 中国科学院沈阳自动化研究所 | Intelligent detection system for highway vehicle violation behaviors |
| CN112116813A (en) * | 2020-11-20 | 2020-12-22 | 深圳市城市交通规划设计研究中心股份有限公司 | State recognition model establishing method, road traffic state recognition method and road traffic state recognition device |
| CN112149595A (en) * | 2020-09-29 | 2020-12-29 | 爱动超越人工智能科技（北京）有限责任公司 | Method for detecting lane line and vehicle violation by using unmanned aerial vehicle |
| CN112164227A (en) * | 2020-08-26 | 2021-01-01 | 深圳奇迹智慧网络有限公司 | Illegal parking vehicle warning method and device, computer equipment and storage medium |
| CN112329747A (en) * | 2021-01-04 | 2021-02-05 | 湖南大学 | Vehicle parameter detection method based on video identification and deep learning and related device |
| WO2021022759A1 (en) * | 2019-08-08 | 2021-02-11 | Zhejiang Dahua Technology Co., Ltd. | Systems and methods for traffic violation detection |
| CN112365741A (en) * | 2020-10-23 | 2021-02-12 | 淮阴工学院 | Safety early warning method and system based on multilane vehicle distance detection |
| CN112365549A (en) * | 2021-01-12 | 2021-02-12 | 腾讯科技（深圳）有限公司 | Attitude correction method and device for vehicle-mounted camera, storage medium and electronic device |
| CN113593235A (en) * | 2021-08-03 | 2021-11-02 | 安徽迪卡智能控制技术有限公司 | Traffic monitoring system based on internet of things technology |
| CN113593257A (en) * | 2021-07-19 | 2021-11-02 | 山东仁智信息工程有限公司 | Speed limiting method for vehicles on highway |
| CN113903180A (en) * | 2021-11-17 | 2022-01-07 | 四川九通智路科技有限公司 | Method and system for detecting vehicle overspeed on expressway |
| CN114067564A (en) * | 2021-11-15 | 2022-02-18 | 武汉理工大学 | Traffic condition comprehensive monitoring method based on YOLO |
| CN114155710A (en) * | 2021-11-29 | 2022-03-08 | 武汉中交交通工程有限责任公司 | Underground road confluence section guide control system and control method |
| CN114170580A (en) * | 2021-12-01 | 2022-03-11 | 连云港杰瑞电子有限公司 | Highway-oriented abnormal event detection method |
| US20220092797A1 (en) * | 2019-07-08 | 2022-03-24 | Zhongyuan University Of Technology | Intelligent Vehicle Trajectory Measurement Method Based on Binocular Stereo Vision System |
| CN114332385A (en) * | 2021-11-23 | 2022-04-12 | 南京国图信息产业有限公司 | Monocular camera target detection and spatial positioning method based on three-dimensional virtual geographic scene |
| CN114399645A (en) * | 2021-12-20 | 2022-04-26 | 西安电子科技大学 | Multi-mode data expansion method, system, medium, computer equipment and terminal |
| CN114495512A (en) * | 2021-12-24 | 2022-05-13 | 武汉万集光电技术有限公司 | Vehicle information detection method and system, electronic device and readable storage medium |
| CN114530042A (en) * | 2021-12-31 | 2022-05-24 | 威海南海数字产业研究院有限公司 | Urban traffic brain monitoring system based on internet of things technology |
-
2022
- 2022-06-14 CN CN202210666406.8A patent/CN114758511B/en active Active
Patent Citations (36)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106205134A (en) * | 2015-04-29 | 2016-12-07 | 上海宝康电子控制工程有限公司 | The comprehensive vehicle detecting system of highway |
| CN105702048A (en) * | 2016-03-23 | 2016-06-22 | 武汉理工大学 | Automobile-data-recorder-based illegal lane occupation identification system and method for automobile on highway |
| CN106530746A (en) * | 2016-10-28 | 2017-03-22 | 江苏信息职业技术学院 | Vehicle speed detection apparatus and detection method based on video technology |
| US20180365858A1 (en) * | 2017-06-14 | 2018-12-20 | Hyundai Mobis Co., Ltd. | Calibration method and apparatus |
| CN107958590A (en) * | 2017-11-20 | 2018-04-24 | 北京工业大学 | A kind of traffic offence crimping seizes lanes behavioral value acquisition system and method |
| CN110178167A (en) * | 2018-06-27 | 2019-08-27 | 潍坊学院 | Crossing video frequency identifying method violating the regulations based on video camera collaboration relay |
| CN109102702A (en) * | 2018-08-24 | 2018-12-28 | 南京理工大学 | Vehicle speed measuring method based on video encoder server and Radar Signal Fusion |
| JP2020042712A (en) * | 2018-09-13 | 2020-03-19 | 株式会社Ｊｖｃケンウッド | Video recording control device, video recording system, video recording method, and video recording control program |
| CN109615870A (en) * | 2018-12-29 | 2019-04-12 | 南京慧尔视智能科技有限公司 | A kind of traffic detection system based on millimetre-wave radar and video |
| CN110399766A (en) * | 2019-01-28 | 2019-11-01 | 浙江浩腾电子科技股份有限公司 | Smoking testing and analysis system based on deep learning |
| WO2020180620A1 (en) * | 2019-03-01 | 2020-09-10 | Amazon Technologies, Inc. | Identifying parking location using single camera reverse projection |
| US20220092797A1 (en) * | 2019-07-08 | 2022-03-24 | Zhongyuan University Of Technology | Intelligent Vehicle Trajectory Measurement Method Based on Binocular Stereo Vision System |
| WO2021022759A1 (en) * | 2019-08-08 | 2021-02-11 | Zhejiang Dahua Technology Co., Ltd. | Systems and methods for traffic violation detection |
| CN110909699A (en) * | 2019-11-28 | 2020-03-24 | 北京以萨技术股份有限公司 | Video vehicle non-guide driving detection method and device and readable storage medium |
| CN111009131A (en) * | 2019-12-05 | 2020-04-14 | 成都思晗科技股份有限公司 | High-order video intelligence parking system based on image recognition |
| CN111402593A (en) * | 2020-03-02 | 2020-07-10 | 哈尔滨工业大学 | Video traffic parameter acquisition method based on polynomial fitting |
| CN112102609A (en) * | 2020-03-10 | 2020-12-18 | 中国科学院沈阳自动化研究所 | Intelligent detection system for highway vehicle violation behaviors |
| CN111915883A (en) * | 2020-06-17 | 2020-11-10 | 西安交通大学 | Road traffic condition detection method based on vehicle-mounted camera shooting |
| CN111753797A (en) * | 2020-07-02 | 2020-10-09 | 浙江工业大学 | Vehicle speed measuring method based on video analysis |
| CN112037159A (en) * | 2020-07-29 | 2020-12-04 | 长安大学 | Cross-camera road space fusion and vehicle target detection tracking method and system |
| CN112164227A (en) * | 2020-08-26 | 2021-01-01 | 深圳奇迹智慧网络有限公司 | Illegal parking vehicle warning method and device, computer equipment and storage medium |
| CN112149595A (en) * | 2020-09-29 | 2020-12-29 | 爱动超越人工智能科技（北京）有限责任公司 | Method for detecting lane line and vehicle violation by using unmanned aerial vehicle |
| CN112365741A (en) * | 2020-10-23 | 2021-02-12 | 淮阴工学院 | Safety early warning method and system based on multilane vehicle distance detection |
| CN112116813A (en) * | 2020-11-20 | 2020-12-22 | 深圳市城市交通规划设计研究中心股份有限公司 | State recognition model establishing method, road traffic state recognition method and road traffic state recognition device |
| CN112329747A (en) * | 2021-01-04 | 2021-02-05 | 湖南大学 | Vehicle parameter detection method based on video identification and deep learning and related device |
| CN112365549A (en) * | 2021-01-12 | 2021-02-12 | 腾讯科技（深圳）有限公司 | Attitude correction method and device for vehicle-mounted camera, storage medium and electronic device |
| CN113593257A (en) * | 2021-07-19 | 2021-11-02 | 山东仁智信息工程有限公司 | Speed limiting method for vehicles on highway |
| CN113593235A (en) * | 2021-08-03 | 2021-11-02 | 安徽迪卡智能控制技术有限公司 | Traffic monitoring system based on internet of things technology |
| CN114067564A (en) * | 2021-11-15 | 2022-02-18 | 武汉理工大学 | Traffic condition comprehensive monitoring method based on YOLO |
| CN113903180A (en) * | 2021-11-17 | 2022-01-07 | 四川九通智路科技有限公司 | Method and system for detecting vehicle overspeed on expressway |
| CN114332385A (en) * | 2021-11-23 | 2022-04-12 | 南京国图信息产业有限公司 | Monocular camera target detection and spatial positioning method based on three-dimensional virtual geographic scene |
| CN114155710A (en) * | 2021-11-29 | 2022-03-08 | 武汉中交交通工程有限责任公司 | Underground road confluence section guide control system and control method |
| CN114170580A (en) * | 2021-12-01 | 2022-03-11 | 连云港杰瑞电子有限公司 | Highway-oriented abnormal event detection method |
| CN114399645A (en) * | 2021-12-20 | 2022-04-26 | 西安电子科技大学 | Multi-mode data expansion method, system, medium, computer equipment and terminal |
| CN114495512A (en) * | 2021-12-24 | 2022-05-13 | 武汉万集光电技术有限公司 | Vehicle information detection method and system, electronic device and readable storage medium |
| CN114530042A (en) * | 2021-12-31 | 2022-05-24 | 威海南海数字产业研究院有限公司 | Urban traffic brain monitoring system based on internet of things technology |
Non-Patent Citations (5)
| Title |
|---|
| GUO-HANG SHAN: "Carcorder Camera Calibration method of external parameters based on lane line", 《2020 7TH INTERNATIONAL CONFERENCE ON INFORMATION SCIENCE AND CONTROL ENGINEERING (ICISCE)》 * |
| 侯雨彤: "基于无人机航拍视频的车辆轨迹提取方法研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
| 孙宁: "一种提高视频车速检测精度的算法分析和实现", 《合肥工业大学学报(自然科学版)》 * |
| 孟乔: "面向复杂场景的车辆检测跟踪及行为分析关键技术研究", 《中国博士学位论文全文数据库 工程科技Ⅱ辑》 * |
| 莫德强: "基于无人机平台的道路车辆违章超速行为检测算法研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114758511B (en) | 2022-11-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111368687B (en) | Sidewalk vehicle illegal parking detection method based on target detection and semantic segmentation | |
| CN110263706B (en) | Method for detecting and identifying dynamic target of vehicle-mounted video in haze weather | |
| US11380104B2 (en) | Method and device for detecting illegal parking, and electronic device | |
| CN110689724B (en) | Automatic motor vehicle zebra crossing present pedestrian auditing method based on deep learning | |
| US11790499B2 (en) | Certificate image extraction method and terminal device | |
| CN111967396A (en) | Processing method, device and equipment for obstacle detection and storage medium | |
| CN114170580A (en) | Highway-oriented abnormal event detection method | |
| CN101320048A (en) | Multi-charge coupled device image sensor large visual field vehicle speed measuring device arranged in sector form | |
| CN114898243A (en) | Traffic scene analysis method and device based on video stream | |
| CN114758511B (en) | Sports car overspeed detection system, method, electronic equipment and storage medium | |
| CN117152513A (en) | Vehicle boundary positioning method for night scene | |
| CN115601717B (en) | Deep learning-based traffic offence behavior classification detection method and SoC chip | |
| CN111160282A (en) | Traffic light detection method based on binary Yolov3 network | |
| CN116597270A (en) | Road damage target detection method based on attention mechanism integrated learning network | |
| CN113901961B (en) | Parking space detection method, device, equipment and storage medium | |
| CN115909241A (en) | Lane line detection method, system, electronic device and storage medium | |
| CN113902047B (en) | Image element matching method, device, equipment and storage medium | |
| CN116343136A (en) | Road surface casting detection method based on expressway monitoring video | |
| CN112950954B (en) | Intelligent parking license plate recognition method based on high-position camera | |
| CN115909140A (en) | Video target segmentation method and system based on high-order video monitoring | |
| CN115546667A (en) | Real-time lane line detection method for unmanned aerial vehicle scene | |
| CN101577052B (en) | Device and method for detecting vehicles by overlooking | |
| CN115393822A (en) | Method and equipment for detecting obstacle in driving in foggy weather | |
| CN113343817A (en) | Unmanned vehicle path detection method and device for target area and medium | |
| TWI823819B (en) | Driving assistance system and driving assistance computation method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |