Research On Vehicle Violation Pressure Line Detection Algorithm Based On Deep Learning

In recent years,the number of cars has been rapidly increasing in our lives,and traffic accidents caused by cars is more and more common,which mainly caused by the bad habits and illegal operations of drivers,among them the vehicle crimping is the most common illegal driving operation,accounting for the highest proportion of traffic accident mortality.Surveillance cameras are been deemed as an important tool to monitor traffic in traffic management system,which can cause a certain deterrent to the driver.However,this approach is unable to provide a sufficient number of cameras to achieve full coverage.Therefore,this paper aims to uses vehicle cameras with extensive coverage to monitor traffic and the experiment is conducted based on deep learning,which extracts useful information from the vehicle image,builds a vehicle pressure line detection algorithm model,and studies the vehicle pressure line in the vehicle image.The main research contents of this article are as follows:(1)Considering the deep learning detection algorithm need large amount of data,this paper builds the data set required for the experiment according to the needs and pre-annotating the constructed data,which save a lot of manual time,and then use the annotation tool to modify it.(2)Since the image quality is affected by factors such as illumination and weather during the collection process,the paper chooses median filtering to weaken the image noise after analyzing a large number of image preprocessing methods;HSV color image histogram equalization is used to enhance the image contrast.(3)Google Net-FCN semantic segmentation network is conducted to extract lane line and wheel-line line information,perform morphological filtering on the segmentation information and connected area threshold method to improve the segmentation effect,and the model evaluates the MIo U value on the test set to 66.2%.This paper improve the Mobile Net-SSD network to modify the prediction branch and anchor value according to the specific data set,and perform feature fusion.Therefore,the accuracy of the network performance on the data set is 95.7%,and the recall rate is 94.6%.(4)This paper establish a pressure line algorithm model,and use the information extracted by the deep learning model to judge the line pressure of the vehicle.According to the position of the intersection of the lane line and the wheel-line line,a threshold is set to determine whether the line is pressed,and the best th reshold is found according to the evaluation results of the different thresholds on the line pressure data.Finally,the performance of the algorithm is analyzed,which shows that the average time of the algorithm is 67.8ms and the accuracy rate is 96.6%.It is proved that the target vehicle pressure line detection algorithm in this experiment has good accuracy and robustness,and it can also meet the real-time requirements.