Motorcycle helmet wearing detection method based on YOLOv4
A detection method and motorcycle technology, applied in neural learning methods, instruments, biological neural network models, etc., can solve problems such as poor application of helmet detection, difficulty in distinguishing target categories, multi-detection and false detection, etc.
Inactive Publication Date: 2021-04-13
HARBIN UNIV OF SCI & TECH
View PDF5 Cites 1 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0004] Most of the current motorcycle helmet wearing detection algorithms are based on specific detection directions, such as detecting the front view angle or the rear view angle of the motorcycle. Although this can accurately detect the helmet wearing situation of the motorcycle rider and the motorcycle license plate information, However, due to the mutual occlusion between the riders, this method cannot be well applied to the multi-rider helmet detection situation.
[0005] The currently known motorcycle helmet wearing detection methods have multiple detections and false detections when detecting targets, especially in the case of dense vehicles and occlusions, it is even more difficult to distinguish target categories
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0026] The present invention will be further described below in conjunction with accompanying drawing.
[0027] like figure 1 Shown: the present invention comprises the steps:
[0028] Step 1: Class coding for motorcycle and rider areas;
[0029] Step 2: Improve the non-maximum suppression algorithm (Non-Maximum Suppression, NMS) of YOLOv4, and train the YOLOv4 detection network for the motorcycle and rider areas;
[0030] Step 3: By inputting traffic monitoring video frames, passing through the motorcycle and rider area detection network, when a single rider and two riders are detected, use a rectangular frame to mark the target area and record the number of riders, extract and send Enter the next module; when more than two riders are detected, directly output the target area and mark it as overload;
[0031] Step 4: Perform helmet ROI on the extracted motorcycle and rider area, and the ROI area will be directly sent to the next module for helmet detection;
[0032] Step ...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention relates to a motorcycle helmet wearing detection method based on an improved YOLOv4 algorithm. The method provided by the invention is divided into three modules: a motorcycle and rider area detection module, a helmet ROI module and a helmet detection module. For motorcycle and rider region detection, a YOLOv4 target detection network is combined with a multi-rider category code to carry out region detection and extraction; for the helmet ROI module, data set statistical information is used to determine an ROI region; for the helmet detection module, we also train a YOLOv4 network to specially perform helmet detection. The motorcycle helmet wearing detection method effectively solves the problems of false detection and multi-detection of the target, increases the detection support for multiple riders, and improves the motorcycle helmet wearing detection speed and precision at the same time.
Description
technical field [0001] The invention belongs to the technical field of image processing and computer vision, relates to target detection technology, in particular to a method for detecting helmet wearing of motorcycle riders, and is mainly used in the field of road traffic safety. Background technique [0002] The increase of vehicles not only promotes the progress of human civilization, but also increases the frequency of road safety accidents. In most developing countries, two-wheelers share the motorway with larger trucks, buses and cars, so the risk of accidents is high. According to the survey, helmets are the only safety equipment for two-wheeled motorcycles. Using motorcycle helmets according to the requirements can reduce the possibility of motorcyclists suffering fatal injuries in road traffic accidents by 42%. In order to ensure the safety of cyclists on the road, it is necessary to detect whether motorcycle riders wear helmets on the road. The growth of populatio...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More IPC IPC(8): G06K9/00G06K9/32G06N3/04G06N3/08
CPCG06N3/08G06V20/46G06V20/41G06V10/25G06N3/045
Inventor 李启瑞贾伟楠吕衍河尹芳
Owner HARBIN UNIV OF SCI & TECH