Vehicle detection method and system based on monocular vision and deep learning

Abstract

The invention provides a vehicle detection method and system based on monocular vision and deep learning, and the method comprises the steps: obtaining expressway video frame image data, employing an evolutionary algorithm to optimize camera parameters of back projection from a two-dimensional space to a three-dimensional space, carrying out the camera calibration, and obtaining the calibrated camera parameters; a neural network is adopted to carry out vehicle detection, a vehicle type and a two-dimensional bounding box are obtained, the calibrated camera parameters are adopted to track back foot points of the vehicle two-dimensional bounding box to a three-dimensional space, and vehicle coordinate points in a world coordinate system are obtained; calculating a vehicle distance by adopting Euclidean distance according to a vehicle coordinate point under a world coordinate system and a zero point coordinate determined by a camera, calculating a vehicle speed according to the vehicle distance based on a sliding time window algorithm, judging whether a speed variance is smaller than a given threshold value or not, and judging the vehicle with too low speed as a stopped vehicle; rapid detection of the vehicle distance and the vehicle speed is realized, and the real-time performance of detection is improved.

Description

technical field [0001] The invention relates to the technical field of vehicle detection, in particular to a vehicle detection method and system based on monocular vision and deep learning. Background technique [0002] The statements in this section merely provide background art related to the present invention and do not necessarily constitute prior art. [0003] The safety of highway vehicles has always been the focus of social attention. The initial highway monitoring system required personnel to view the video in the monitoring background. After a traffic accident occurred, the personnel reported to the monitoring background to learn of the accident and then make emergency treatment; Prediction and early warning of primary accidents and secondary accidents cannot avoid the occurrence of primary accidents and secondary accidents. With the rise and development of new technologies such as big data and artificial intelligence, intelligent traffic management systems have be...

AI Technical Summary

Problems solved by technology

[0003] The safety of highway vehicles has always been the focus of social attention. The initial monitoring system of the highway required personnel to view the video in the monitoring background. Make predictions and early warnings for primary accidents and secondary accidents, and cannot avoid the occurrence of primary accidents and secondary accidents

[0005] The inventors found that with the increase of video image quality, the existing 5G network infrastructure does not cover all, and the 4G network bandwidth cannot meet the real real-time transmission of high-quality video images. Carry out further computer vision processing to measure vehicle speed and distance, and it is even more impossible to guarantee the real-time performance of vehicle speed measurement and distance measurement

Images