CN110929676A - Deep learning-based real-time detection method for illegal turning around - Google Patents
Deep learning-based real-time detection method for illegal turning around Download PDFInfo
- Publication number
- CN110929676A CN110929676A CN201911228884.5A CN201911228884A CN110929676A CN 110929676 A CN110929676 A CN 110929676A CN 201911228884 A CN201911228884 A CN 201911228884A CN 110929676 A CN110929676 A CN 110929676A
- Authority
- CN
- China
- Prior art keywords
- tracking
- target
- turning around
- time
- violation
- 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
- 238000000034 method Methods 0.000 title claims abstract description 37
- 238000013135 deep learning Methods 0.000 title claims abstract description 30
- 238000011897 real-time detection Methods 0.000 title claims abstract description 13
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 13
- 238000001514 detection method Methods 0.000 claims description 35
- 230000006399 behavior Effects 0.000 claims description 6
- 239000002131 composite material Substances 0.000 claims description 4
- 230000007704 transition Effects 0.000 claims description 3
- 238000004458 analytical method Methods 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 4
- 230000008859 change Effects 0.000 abstract description 3
- 230000007613 environmental effect Effects 0.000 abstract description 2
- 230000033228 biological regulation Effects 0.000 description 5
- 230000028838 turning behavior Effects 0.000 description 5
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000006698 induction Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 206010039203 Road traffic accident Diseases 0.000 description 1
- 238000007792 addition Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
- G06V20/54—Surveillance or monitoring of activities, e.g. for recognising suspicious objects of traffic, e.g. cars on the road, trains or boats
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
- G06T7/248—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments involving reference images or patches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30232—Surveillance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30236—Traffic on road, railway or crossing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/08—Detecting or categorising vehicles
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a deep learning-based real-time detection method for illegal turning around, which comprises the following steps: 1) setting a camera preset position and calibrating a camera; 2) initializing a convolutional neural network model; 3) acquiring a real-time video stream; 4) checking the working state of the camera; 5) detecting a vehicle target by using a convolutional neural network model; 6) tracking a vehicle target; 7) converting the state of a vehicle tracking target; 8) analyzing the turning around of the violation; 9) and reporting the violation turning around. The invention provides a method for detecting the illegal turning around in real time based on deep learning, which has stronger robustness to environmental change, realizes the real-time detection effect and higher illegal turning around identification precision, simplifies the evidence obtaining work and greatly reduces the cost of human resources.
Description
Technical Field
The invention relates to the technical field of deep learning, in particular to a method for detecting violation turning around in real time based on deep learning.
Background
In recent years, with the popularization of automobiles, the pace of life of people is accelerated, the awareness of traffic safety is deficient, and illegal turning-around behaviors are frequently generated. The illegal turning around is mainly realized by forcibly turning around at intersections where turning around is not allowed. The behavior often causes traffic jam easily, affects the traveling efficiency of people, even more causes serious traffic accidents, and seriously harms the traveling safety of people. Therefore, it is important to accurately detect the illegal turning behavior of the vehicle in real time.
At present, the mode of detecting the violation turning around mainly comprises three modes, the first mode is manual detection of the violation turning around behavior, such as on-site law enforcement of traffic management personnel or manual capturing of the violation turning around behavior by utilizing video monitoring, but the manual mode cannot ensure all-weather inspection and supervision, and has complex work and high labor cost; the second is that sensors such as ground induction lines are used for detecting the illegal turning behavior, but the construction of the ground induction lines needs to damage the road surface, and the construction is complex, easy to damage and difficult to repair; the third is that the computer vision technology is used for realizing automatic detection of illegal turning behavior, at present, most of the traditional methods are used for vehicle detection, the use limitation exists, the phenomena of rain and fog, camera shake, light change and the like can affect the scene, the detection effect is further directly affected, and the robustness of the algorithm is poor.
Disclosure of Invention
In order to overcome the defects of low detection precision and low detection speed in the prior art, the invention provides a method for detecting the violation turning around in real time based on deep learning, which uses the CNN characteristics of a deep convolutional neural network to detect a vehicle target and integrates the video time-space continuity to accurately and quickly detect the violation turning around.
In order to realize the invention, the technical scheme is as follows:
a violation turning around real-time detection method based on deep learning is characterized by comprising the following steps:
1) setting a camera preset position and calibrating a camera;
2) initializing a convolutional neural network model;
3) acquiring a real-time video stream;
4) checking the working state of the camera;
5) detecting a vehicle target by using a convolutional neural network model;
6) tracking a vehicle target;
7) converting the state of a vehicle tracking target;
8) analyzing the turning around of the violation;
9) and reporting the violation turning around.
The real-time detection method for the violation turning around based on the deep learning is characterized in that in the step 1), the preset position of the camera is a fixed position where the camera is located when the violation turning around detection is carried out.
The method for detecting the illegal turning around in real time based on the deep learning is characterized in that in the step 1), the camera preset position is set to be a position for adjusting the camera to be suitable for detecting the illegal turning around, and the current camera position is set to be the preset position.
The real-time detection method for the illegal turning around based on the deep learning is characterized in that in the step 1), the camera is calibrated by intercepting a frame of image of a camera video stream, and calibrating a lane line, a passing detection area, a illegal turning around detection area, an illegal turning around starting line, an illegal turning around intermediate line and an illegal turning around ending line is carried out on the image.
The method for detecting the violation turning around in real time based on deep learning is characterized in that in the step 2), the convolutional neural network model is initialized to be a convolutional network model which is loaded into a GPU (graphics processing unit) display card and weight parameters are recovered.
The method for detecting the illegal turning around in real time based on deep learning is characterized in that in the step 4), the step of checking the working state of the camera specifically comprises the following steps:
4.1, acquiring the position information of the current camera, comparing the position information with a preset position to determine whether the position information is the same as the preset position, and if so, carrying out illegal turning detection; otherwise, carrying out illegal turning detection;
4.2 calculating the current video frame time T according to the formula (1)curAnd the previous frame time TpreTime interval T ofspaceThe units are seconds; if Tspace≥T0Resetting the target tracking queue; otherwise, carrying out normal illegal turning detection; t is0Represents a time threshold in seconds;
Tspace=Tcur-Tpre(1)
the method for detecting the illegal turning around in real time based on deep learning is characterized in that in the step (6), the vehicle target tracking comprises the following steps:
6.1, matching the vehicle detection result with the tracking target:
6.1.1 calculating IoU sum of width and height errors E of the vehicle object D detected by the current frame and the object Q in the tracking queuew、Eh(ii) a IoU are calculated according to equation (2); wide height error Ew、EhRespectively calculating according to a formula (3) and a formula (4);
wherein D isboxBounding box, Q, of vehicle object D for the current frameTrackingBoxFor tracking real-time boundary box of target Q in queue, ∩ is intersection, ∪ is union, DwAnd DhWidth and height Q of real-time tracking bounding box for vehicle target D of current framewAnd QhRespectively tracking the width and the height of a bounding box of the target Q in the tracking queue in real time;
6.1.2 if IoU is more than or equal to IoUt,Ew≤Ewt,Eh≤EhtIf the vehicle is detected, the D and the Q are considered to be the same vehicle, the D is marked as matched, and the Q is marked as detected; IoU thereint,Ewt,EhtIoU for a preset adjustable valuetHas a value range of 0 to 1.0, EwtAnd EhtThe value of (a) is required to be greater than or equal to 0;
6.2, target tracking and updating:
6.2.1: adding an unmatched vehicle target D in a vehicle passing detection area into a tracking queue, wherein an initial boundary frame and a real-time tracking frame are boundary frames of a current target D, an initial image is a current video frame, and initial time Q isinit_timeThe state quantity is marked as start for the time of the current video frame;
6.2.2: if the detected tracking Q is marked, the tracking mode is IoU tracking, and the boundary box of the target D matched with the real-time tracking box is recorded; if the mark is the detected tracking Q, the tracking mode is KCF tracking, and a KCF algorithm is used for updating the real-time tracking boundary box;
6.3, maintaining a tracking queue:
6.3.1: if the tracking target in the tracking queue is not in the illegal turning detection area, removing the tracking target;
6.3.2: if the tracking target Q is not detected in the current frame and the front E exists in the tracking queuetolerate_frameIf the frame is not updated, the target is considered to be driven away, and the tracking target Q is removed from the tracking queue; wherein Etolerate_frameIs a preset adjustable value with the value range of Etolerate_frame≥1。
The method for detecting the illegal turning around in real time based on deep learning is characterized in that in the step 7), the vehicle tracking target state conversion detection comprises the following steps:
7.1: if the state quantity of the tracking target is start and the center point of the real-time tracking frame drives through the violation turning starting line, the state quantity is going1, and the first image of the tracking target is set as the current video image;
7.2: if the state quantity of the tracking target is going1 and the boundary of the real-time tracking frame passes through the violation turning middle line, the state quantity is going2, and the vehicle close-up image of the tracking target is set as the target real-time tracking frame area of the current video image and can be used for further license plate recognition in the later period and the like;
7.2: if the state quantity of the tracking target is going2 and the center point of the real-time tracking frame drives through the violation turning intermediate line, the state quantity is going3, the tracking mode is set to be IoU tracking, and the second image of the tracking target is set to be the current video image;
7.3: if the state quantity of the tracking target is going3 and the center point of the real-time tracking frame passes through the violation end-of-head ending line, the state quantity becomes end and the third image of the tracking target is set as the current video image.
The real-time detection method for the violation turning around based on the deep learning is characterized in that in the step 8), the violation turning around is analyzed to be a traversal tracking queue, if the state quantity of a tracking target is end, it is judged that the tracking target has the violation turning around behavior, reporting is carried out, and finally the tracking target is removed from the tracking queue.
The real-time detection method for the violation turning around based on the deep learning is characterized in that in the step 8), the violation turning around is reported by drawing the position of the target on three images of the violation turning around target, adding video frame time and location information on the bottom edges of the three images, arranging and combining the three images and a sketch image in a left-right mode to form a four-in-one violation synthetic image, and finally reporting the violation turning around information and the four-in-one violation synthetic image.
Compared with the prior art, the invention has the main beneficial effects that:
the invention provides a method for detecting the violation turning around in real time based on deep learning, which adopts an advanced deep learning technology and a simple and effective state transition discrimination mode, has stronger robustness to environmental change, and realizes real-time detection effect and higher violation turning around recognition precision. The method can effectively monitor the on-site violation turning behavior in all weather by using video monitoring, and meanwhile, the automatic four-in-one violation synthetic graph simplifies the evidence obtaining work, greatly reduces the human resource cost, ensures the traffic safety and improves the traffic efficiency.
Drawings
FIG. 1 is a flow chart of the steps of the method of the present invention;
FIG. 2 is a state transition diagram of the method of the present invention;
FIG. 3 is an example of a calibration chart of the method of the present invention;
fig. 4 is a view of the effect of the violation turning around of the method of the present invention.
In the figure: 1-passing detection area, 2-turning around start line against regulations, 3-lane line, 4-turning around ending line against regulations, 5-turning around middle line against regulations, 6-turning around detection area against regulations, and 7-target vehicle with turning around against regulations.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be further described in detail with reference to the accompanying drawings and examples. It should be understood that the detailed description and specific examples, while indicating the scope of the invention, are intended for purposes of illustration only and are not intended to limit the scope of the invention.
As shown in fig. 1 to 4, the method for detecting a violation turning around in real time based on deep learning provided by this embodiment includes the following steps:
9) and reporting the violation turning around.
S1, setting camera preset positions and calibrating the camera.
Specifically, the camera is adjusted to a proper position for carrying out illegal turning detection, and the current camera position is set as a preset position; then, a frame of image of the camera video stream is intercepted, and the calibration of a passing detection area 1, a violation turning starting line 2, a lane line 3, a violation turning ending line 4, a violation turning intermediate line 5 and a violation turning detection area 6 is carried out on the image.
And S2, initializing a convolutional neural network model.
Specifically, the convolutional network model is loaded into the GPU graphics card and the weight parameters are recovered.
And S3, acquiring the real-time video stream.
S4, checking the working state of the camera.
Specifically, the method comprises the following steps:
s4.1, acquiring the position information of the current camera, comparing the position information with a preset position to determine whether the position information is the same as the preset position, and detecting a congestion area if the position information is the same as the preset position; otherwise, not detecting the congestion area;
s4.2 calculating the current video frame time T according to the formula (1)curAnd the previous frame time TpreTime interval T ofspaceThe units are seconds; if Tspace≥T0Resetting the target tracking queue; otherwise, carrying out normal illegal turning detection; t is0Representing a time threshold in seconds.
Tspace=Tcur-Tpre(1)
And S5, detecting the vehicle target by using the convolutional neural network model.
And S6, tracking the vehicle target.
Specifically, the method comprises the following steps:
s6.1, matching a vehicle detection result with a tracking target:
s6.1.1 calculating IoU and width-height error E of vehicle object D detected by current frame and object Q in tracking queuew、Eh(ii) a IoU are calculated according to equation (2); wide height error Ew、EhRespectively calculating according to a formula (3) and a formula (4);
wherein D isboxBounding box, Q, of vehicle object D for the current frameTrackingBoxFor tracking real-time boundary box of target Q in queue, ∩ is intersection, ∪ is union, DwAnd DhWidth and height Q of real-time tracking bounding box for vehicle target D of current framewAnd QhRespectively tracking the width and the height of a bounding box of the target Q in the tracking queue in real time;
s6.1.2 if IoU is not less than IoUt,Ew≤Ewt,Eh≤EhtIf the vehicle is detected, the D and the Q are considered to be the same vehicle, the D is marked as matched, and the Q is marked as detected; IoU thereint,Ewt,EhtIoU for a preset adjustable valuetHas a value range of 0 to 1.0, EwtAnd EhtThe value of (a) is required to be greater than or equal to 0;
s6.2, target tracking and updating:
s6.2.1: adding an unmatched vehicle target D in a vehicle passing detection area into a tracking queue, wherein an initial boundary frame and a real-time tracking frame are boundary frames of a current target D, an initial image is a current video frame, and initial time Q isinit_timeThe state quantity is marked as start for the time of the current video frame;
s6.2.2: if the detected tracking Q is marked, the tracking mode is IoU tracking, and the boundary box of the target D matched with the real-time tracking box is recorded; if the mark is the detected tracking Q, the tracking mode is KCF tracking, and a KCF algorithm is used for updating the real-time tracking boundary box;
s6.3, maintaining a tracking queue:
s6.3.1: if the tracking target in the tracking queue is not in the illegal turning detection area, removing the tracking target;
s6.3.2: if the tracking target Q is not detected in the current frame and the front E exists in the tracking queuetolerate_frameIf the frame is not updated, the target is considered to be driven away, and the tracking target Q is removed from the tracking queue; wherein Etolerate_frameIs a preset adjustable value with the value range of Etolerate_frame≥1;
S7, vehicle tracking target state conversion
Specifically, the method comprises the following steps:
s7.1: if the state quantity of the tracking target is start and the center point of the real-time tracking frame drives through the violation turning starting line, the state quantity is going1, and the first image of the tracking target is set as the current video image;
s7.2: if the state quantity of the tracking target is going1 and the boundary of the real-time tracking frame passes through the violation turning middle line, the state quantity is going2, and the vehicle close-up image of the tracking target is set as the target real-time tracking frame area of the current video image and can be used for further license plate recognition in the later period and the like;
s7.2: if the state quantity of the tracking target is going2 and the center point of the real-time tracking frame drives through the violation turning intermediate line, the state quantity is going3, the tracking mode is set to be IoU tracking, and the second image of the tracking target is set to be the current video image;
s7.3: if the state quantity of the tracking target is going3 and the center point of the real-time tracking frame passes through the violation end-falling ending line, the state quantity is changed to end, and the third image of the tracking target is set as the current video image;
s8, analysis of illegal turning around
Specifically, traversing the tracking queue, if the state quantity of the tracking target is end, judging that the tracking target has a violation turning behavior, reporting, and finally removing the tracking target from the tracking queue;
s8, reporting the illegal turning around
Specifically, the position of the target, namely the target vehicle 7 with the violation turning around is drawn on the three images of the violation turning around target, the three images and a sketch map are arranged and combined in pairs in a left-right direction to form a four-in-one violation composite map after video frame time and location information is added to the bottom edges of the three images, and finally the violation turning around information and the four-in-one violation composite map are reported.
The above-mentioned embodiments are intended to illustrate the technical solutions and advantages of the present invention, and it should be understood that the above-mentioned embodiments are only the most preferred embodiments of the present invention, and are not intended to limit the present invention, and any modifications, additions, equivalents, etc. made within the scope of the principles of the present invention should be included in the scope of the present invention.
Claims (10)
1. A violation turning around real-time detection method based on deep learning is characterized by comprising the following steps:
1) setting a camera preset position and calibrating a camera;
2) initializing a convolutional neural network model;
3) acquiring a real-time video stream;
4) checking the working state of the camera;
5) detecting a vehicle target by using a convolutional neural network model;
6) tracking a vehicle target;
7) converting the state of a vehicle tracking target;
8) analyzing the turning around of the violation;
9) and reporting the violation turning around.
2. The method for detecting the illegal turning around based on the deep learning as claimed in claim 1, wherein in the step 1), the preset position of the camera is a fixed position where the camera is located when the illegal turning around detection is carried out.
3. The method for detecting the illegal turning around based on the deep learning as claimed in claim 1, wherein in the step 1), the camera preset position is set as a position for adjusting the camera to be suitable for illegal turning around detection, and the current camera position is set as the preset position.
4. The method for detecting the violation turning around in real time based on the deep learning as claimed in claim 1, wherein in the step 1), the camera is calibrated to intercept one frame of image of the camera video stream, and the calibration of a lane line, a passing vehicle detection area, a violation turning around starting line, a violation turning around middle line and a violation turning around ending line is performed on the image.
5. The deep learning-based violation turning around real-time detection method according to claim 1, wherein in the step 2), the convolutional neural network model is tiny-YOLO-v3, and the convolutional neural network model is initialized to be loaded into a GPU graphics card and weight parameters are restored.
6. The method for detecting the illegal turning around in real time based on the deep learning as claimed in claim 1, wherein in the step 4), the step of checking the working state of the camera specifically comprises the following steps:
4.1, acquiring the position information of the current camera, comparing the position information with a preset position to determine whether the position information is the same as the preset position, and if so, carrying out illegal turning detection; otherwise, carrying out illegal turning detection;
4.2 calculating the current video frame time T according to the formula (1)curAnd the previous frame time TpreTime interval T ofspaceThe units are seconds; if Tspace≥T0Resetting the target tracking queue; otherwise, carrying out normal illegal turning detection; t is0Represents a time threshold in seconds;
Tspace=Tcur-Tpre(1)
7. the method for detecting the illegal turning around in real time based on the deep learning as claimed in claim 1, wherein in the step (6), the step of tracking the vehicle target comprises the following steps:
6.1, matching the vehicle detection result with the tracking target:
6.1.1 calculating IoU sum of width and height errors E of the vehicle object D detected by the current frame and the object Q in the tracking queuew、Eh(ii) a IoU are calculated according to equation (2); wide height error Ew、EhRespectively calculating according to a formula (3) and a formula (4);
wherein D isboxBounding box, Q, of vehicle object D for the current frameTrackingBoxFor tracking real-time boundary box of target Q in queue, ∩ is intersection, ∪ is union, DwAnd DhWidth and height Q of real-time tracking bounding box for vehicle target D of current framewAnd QhRespectively tracking the width and the height of a bounding box of the target Q in the tracking queue in real time;
6.1.2 if IoU is more than or equal to IoUt,Ew≤Ewt,Eh≤EhtIf the vehicle is detected, the D and the Q are considered to be the same vehicle, the D is marked as matched, and the Q is marked as detected; IoU thereint,Ewt,EhtIoU for a preset adjustable valuetHas a value range of 0 to 1.0, EwtAnd EhtThe value of (a) is required to be greater than or equal to 0;
6.2, target tracking and updating:
6.2.1: adding an unmatched vehicle target D in a vehicle passing detection area into a tracking queue, wherein an initial boundary frame and a real-time tracking frame are boundary frames of a current target D, an initial image is a current video frame, and initial time Q isinit_timeThe state quantity is marked as start for the time of the current video frame;
6.2.2: if the detected tracking Q is marked, the tracking mode is IoU tracking, and the boundary box of the target D matched with the real-time tracking box is recorded; if the mark is the detected tracking Q, the tracking mode is KCF tracking, and a KCF algorithm is used for updating the real-time tracking boundary box;
6.3, maintaining a tracking queue:
6.3.1: if the tracking target in the tracking queue is not in the illegal turning detection area, removing the tracking target;
6.3.2: if the tracking target Q is not detected in the current frame and the front E exists in the tracking queuetolerate_frameIf the frame is not updated, the target is considered to be driven away, and the tracking target Q is removed from the tracking queue; wherein Etolerate_frameIs a preset adjustable value with the value range of Etolerate_frame≥1。
8. The method for detecting the illegal turning around in real time based on the deep learning as claimed in claim 1, wherein in the step 7), the detection of the state transition of the vehicle tracking target comprises the following steps:
7.1: if the state quantity of the tracking target is start and the center point of the real-time tracking frame drives through the violation turning starting line, the state quantity is going1, and the first image of the tracking target is set as the current video image;
7.2: if the state quantity of the tracking target is going1 and the boundary of the real-time tracking frame passes through the violation turning middle line, the state quantity is going2, and the vehicle close-up image of the tracking target is set as the target real-time tracking frame area of the current video image and can be used for further license plate recognition in the later period and the like;
7.2: if the state quantity of the tracking target is going2 and the center point of the real-time tracking frame drives through the violation turning intermediate line, the state quantity is going3, the tracking mode is set to be IoU tracking, and the second image of the tracking target is set to be the current video image;
7.3: if the state quantity of the tracking target is going3 and the center point of the real-time tracking frame passes through the violation end-of-head ending line, the state quantity becomes end and the third image of the tracking target is set as the current video image.
9. The deep learning-based real-time detection method for the turning around of the violation, as set forth in claim 1, wherein in step 8), the analysis of the turning around of the violation is a traversal tracking queue, if the state quantity of the tracking target is end, it is determined that the tracking target has a turning around behavior against the violation, reporting is performed, and finally the tracking target is removed from the tracking queue.
10. The method for detecting the violation turning around in real time based on the deep learning of claim 1, wherein in the step 8), the violation turning around reporting is to draw the position of the target on three images of the violation turning around target, add video frame time and location information on the bottom edge of the three images, arrange and combine the three images and a sketch map into a four-in-one violation composite map in a left-right arrangement mode, and report the violation turning around information and the four-in-one violation composite map.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911228884.5A CN110929676A (en) | 2019-12-04 | 2019-12-04 | Deep learning-based real-time detection method for illegal turning around |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911228884.5A CN110929676A (en) | 2019-12-04 | 2019-12-04 | Deep learning-based real-time detection method for illegal turning around |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110929676A true CN110929676A (en) | 2020-03-27 |
Family
ID=69856759
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911228884.5A Pending CN110929676A (en) | 2019-12-04 | 2019-12-04 | Deep learning-based real-time detection method for illegal turning around |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110929676A (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111627217A (en) * | 2020-05-29 | 2020-09-04 | 济南博观智能科技有限公司 | Method, device and equipment for judging vehicle illegal turning around and storage medium |
CN111932908A (en) * | 2020-08-05 | 2020-11-13 | 浙江工业大学 | Deep learning-based steering ratio and traffic flow statistical method |
CN112528924A (en) * | 2020-12-21 | 2021-03-19 | 上海眼控科技股份有限公司 | Vehicle turning detection method, device, equipment and storage medium |
CN112820115A (en) * | 2021-01-05 | 2021-05-18 | 四川铁投信息技术产业投资有限公司 | Method for identifying steering state of running vehicle |
CN113077635A (en) * | 2021-03-26 | 2021-07-06 | 天津天地伟业智能安全防范科技有限公司 | Traffic violation turning snapshot method and device, electronic equipment and storage medium |
CN113257033A (en) * | 2021-07-01 | 2021-08-13 | 成都宜泊信息科技有限公司 | Parking lot management method and system, storage medium and electronic equipment |
WO2021237749A1 (en) * | 2020-05-29 | 2021-12-02 | Siemens Aktiengesellschaft | Method and apparatus for object tracking and reidentification |
Citations (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR100444762B1 (en) * | 2004-03-10 | 2004-08-18 | 주식회사 비츠로시스 | Lane Violation Detection, Tracking System and the Method |
JP2007316856A (en) * | 2006-05-24 | 2007-12-06 | Sumitomo Electric Ind Ltd | Traveling object detecting device, computer program, and traveling object detection method |
JP2011118485A (en) * | 2009-11-30 | 2011-06-16 | Fujitsu Ten Ltd | U-turn supporting device and method |
CN102622884A (en) * | 2012-03-22 | 2012-08-01 | 杭州电子科技大学 | Vehicle illegal turning behavior detection method based on tracking |
CN102903239A (en) * | 2012-09-27 | 2013-01-30 | 安科智慧城市技术(中国)有限公司 | Method and system for detecting illegal left-and-right steering of vehicle at traffic intersection |
CN103065470A (en) * | 2012-12-18 | 2013-04-24 | 浙江工业大学 | Detection device for behaviors of running red light of vehicle based on machine vision with single eye and multiple detection faces |
CN105574502A (en) * | 2015-12-15 | 2016-05-11 | 中海网络科技股份有限公司 | Automatic detection method for violation behaviors of self-service card sender |
CN105632183A (en) * | 2016-01-27 | 2016-06-01 | 福建工程学院 | Vehicle violation behavior proof method and system thereof |
WO2016113973A1 (en) * | 2015-01-14 | 2016-07-21 | オムロン株式会社 | Traffic violation management system and traffic violation management method |
JP2016133875A (en) * | 2015-01-16 | 2016-07-25 | 住友電工システムソリューション株式会社 | Analyzing apparatus, analyzing method, and analyzing program |
CN106205135A (en) * | 2015-04-30 | 2016-12-07 | 北京文安智能技术股份有限公司 | A kind of detection method of vehicle behavior that turns around violating the regulations, Apparatus and system and a kind of ball machine |
CN106291597A (en) * | 2015-05-26 | 2017-01-04 | 深圳市腾讯计算机系统有限公司 | The monitoring method and apparatus of bus running state |
CN106874863A (en) * | 2017-01-24 | 2017-06-20 | 南京大学 | Vehicle based on depth convolutional neural networks is disobeyed and stops detection method of driving in the wrong direction |
CN107038868A (en) * | 2017-05-08 | 2017-08-11 | 钟辉 | The judgement system violating the regulations that a kind of automobile traffic turns around |
CN107293116A (en) * | 2017-06-27 | 2017-10-24 | 浙江浩腾电子科技股份有限公司 | A kind of traffic incident detecting system based on video analysis |
CN108171980A (en) * | 2018-02-06 | 2018-06-15 | 长沙智能驾驶研究院有限公司 | Break in traffic rules and regulations detection method, system and computer readable storage medium |
CN108847032A (en) * | 2018-08-21 | 2018-11-20 | 北京深瞐科技有限公司 | A kind of traffic violation recognition methods and device |
CN109285355A (en) * | 2018-10-19 | 2019-01-29 | 天津天地人和企业管理咨询有限公司 | A kind of front and back candid photograph traffic cameras system |
WO2019145018A1 (en) * | 2018-01-23 | 2019-08-01 | Siemens Aktiengesellschaft | System, device and method for detecting abnormal traffic events in a geographical location |
CN110287905A (en) * | 2019-06-27 | 2019-09-27 | 浙江工业大学 | A kind of traffic congestion region real-time detection method based on deep learning |
-
2019
- 2019-12-04 CN CN201911228884.5A patent/CN110929676A/en active Pending
Patent Citations (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR100444762B1 (en) * | 2004-03-10 | 2004-08-18 | 주식회사 비츠로시스 | Lane Violation Detection, Tracking System and the Method |
JP2007316856A (en) * | 2006-05-24 | 2007-12-06 | Sumitomo Electric Ind Ltd | Traveling object detecting device, computer program, and traveling object detection method |
JP2011118485A (en) * | 2009-11-30 | 2011-06-16 | Fujitsu Ten Ltd | U-turn supporting device and method |
CN102622884A (en) * | 2012-03-22 | 2012-08-01 | 杭州电子科技大学 | Vehicle illegal turning behavior detection method based on tracking |
CN102903239A (en) * | 2012-09-27 | 2013-01-30 | 安科智慧城市技术(中国)有限公司 | Method and system for detecting illegal left-and-right steering of vehicle at traffic intersection |
CN103065470A (en) * | 2012-12-18 | 2013-04-24 | 浙江工业大学 | Detection device for behaviors of running red light of vehicle based on machine vision with single eye and multiple detection faces |
WO2016113973A1 (en) * | 2015-01-14 | 2016-07-21 | オムロン株式会社 | Traffic violation management system and traffic violation management method |
JP2016133875A (en) * | 2015-01-16 | 2016-07-25 | 住友電工システムソリューション株式会社 | Analyzing apparatus, analyzing method, and analyzing program |
CN106205135A (en) * | 2015-04-30 | 2016-12-07 | 北京文安智能技术股份有限公司 | A kind of detection method of vehicle behavior that turns around violating the regulations, Apparatus and system and a kind of ball machine |
CN106291597A (en) * | 2015-05-26 | 2017-01-04 | 深圳市腾讯计算机系统有限公司 | The monitoring method and apparatus of bus running state |
CN105574502A (en) * | 2015-12-15 | 2016-05-11 | 中海网络科技股份有限公司 | Automatic detection method for violation behaviors of self-service card sender |
CN105632183A (en) * | 2016-01-27 | 2016-06-01 | 福建工程学院 | Vehicle violation behavior proof method and system thereof |
CN106874863A (en) * | 2017-01-24 | 2017-06-20 | 南京大学 | Vehicle based on depth convolutional neural networks is disobeyed and stops detection method of driving in the wrong direction |
CN107038868A (en) * | 2017-05-08 | 2017-08-11 | 钟辉 | The judgement system violating the regulations that a kind of automobile traffic turns around |
CN107293116A (en) * | 2017-06-27 | 2017-10-24 | 浙江浩腾电子科技股份有限公司 | A kind of traffic incident detecting system based on video analysis |
WO2019145018A1 (en) * | 2018-01-23 | 2019-08-01 | Siemens Aktiengesellschaft | System, device and method for detecting abnormal traffic events in a geographical location |
CN108171980A (en) * | 2018-02-06 | 2018-06-15 | 长沙智能驾驶研究院有限公司 | Break in traffic rules and regulations detection method, system and computer readable storage medium |
CN108847032A (en) * | 2018-08-21 | 2018-11-20 | 北京深瞐科技有限公司 | A kind of traffic violation recognition methods and device |
CN109285355A (en) * | 2018-10-19 | 2019-01-29 | 天津天地人和企业管理咨询有限公司 | A kind of front and back candid photograph traffic cameras system |
CN110287905A (en) * | 2019-06-27 | 2019-09-27 | 浙江工业大学 | A kind of traffic congestion region real-time detection method based on deep learning |
Non-Patent Citations (3)
Title |
---|
XIAOLING WANG ET AL: "A video-based traffic violation detection system", 《PROCEEDINGS 2013 INTERNATIONAL CONFERENCE ON MECHATRONIC SCIENCES, ELECTRIC ENGINEERING AND COMPUTER (MEC)》 * |
赵有婷等: "基于视频车辆轨迹模型的交通事件自动检测方法研究", 《 中山大学学报(自然科学版)》 * |
陈伟强: "基于IP Camera的车辆违章行为检测", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111627217A (en) * | 2020-05-29 | 2020-09-04 | 济南博观智能科技有限公司 | Method, device and equipment for judging vehicle illegal turning around and storage medium |
WO2021237749A1 (en) * | 2020-05-29 | 2021-12-02 | Siemens Aktiengesellschaft | Method and apparatus for object tracking and reidentification |
CN111932908A (en) * | 2020-08-05 | 2020-11-13 | 浙江工业大学 | Deep learning-based steering ratio and traffic flow statistical method |
CN111932908B (en) * | 2020-08-05 | 2021-07-23 | 浙江工业大学 | Deep learning-based steering ratio and traffic flow statistical method |
CN112528924A (en) * | 2020-12-21 | 2021-03-19 | 上海眼控科技股份有限公司 | Vehicle turning detection method, device, equipment and storage medium |
CN112820115A (en) * | 2021-01-05 | 2021-05-18 | 四川铁投信息技术产业投资有限公司 | Method for identifying steering state of running vehicle |
CN113077635A (en) * | 2021-03-26 | 2021-07-06 | 天津天地伟业智能安全防范科技有限公司 | Traffic violation turning snapshot method and device, electronic equipment and storage medium |
CN113257033A (en) * | 2021-07-01 | 2021-08-13 | 成都宜泊信息科技有限公司 | Parking lot management method and system, storage medium and electronic equipment |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110929676A (en) | Deep learning-based real-time detection method for illegal turning around | |
CN110287905B (en) | Deep learning-based real-time traffic jam area detection method | |
CN110992693B (en) | Deep learning-based traffic congestion degree multi-dimensional analysis method | |
CN103824452B (en) | A kind of peccancy parking detector based on panoramic vision of lightweight | |
US9704060B2 (en) | Method for detecting traffic violation | |
CN109299674B (en) | Tunnel illegal lane change detection method based on car lamp | |
CN110298307B (en) | Abnormal parking real-time detection method based on deep learning | |
US20110164789A1 (en) | Detection of vehicles in images of a night time scene | |
CN111027447B (en) | Road overflow real-time detection method based on deep learning | |
CN113903008A (en) | Ramp exit vehicle violation identification method based on deep learning and trajectory tracking | |
CN109919062A (en) | A kind of road scene weather recognition methods based on characteristic quantity fusion | |
CN105405297B (en) | A kind of automatic detection method for traffic accident based on monitor video | |
CN111932908A (en) | Deep learning-based steering ratio and traffic flow statistical method | |
CN114663859A (en) | Sensitive and accurate complex road condition lane deviation real-time early warning system | |
CN113468911B (en) | Vehicle-mounted red light running detection method and device, electronic equipment and storage medium | |
CN112329515B (en) | High-point video monitoring congestion event detection method | |
CN116631187B (en) | Intelligent acquisition and analysis system for case on-site investigation information | |
CN110633492A (en) | Lane departure early warning method of Android platform of simulation robot | |
CN114693722B (en) | Vehicle driving behavior detection method, detection device and detection equipment | |
CN105262984A (en) | Detector with fixing device | |
Wang et al. | Detection of lane lines on both sides of road based on monocular camera | |
JP2004030484A (en) | Traffic information providing system | |
JP3771729B2 (en) | Traffic flow measurement system | |
CN112686956A (en) | Method for detecting tilt fault of urban road signal lamp post | |
CN114332159A (en) | Early warning method, system and device for pedestrian illegal crossing guardrail based on target tracking algorithm and storage medium |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200327 |
|
RJ01 | Rejection of invention patent application after publication |