Vehicle intelligent speed measuring method based on binocular stereoscopic vision system

Abstract

The invention provides a vehicle intelligent speed measuring method based on a binocular stereoscopic vision system. The method comprises the following steps: training an SSD neural network by takinga license plate as a detection feature to obtain a license plate identification model; calibrating the binocular stereoscopic vision system to obtain the own parameters and external parameters of twocameras; performing license plate detection on an acquired video frame by using the license plate identification model, and locating the license plate position of a target vehicle; performing featurepoint extraction and stereoscopic matching on the license plate position based on a feature matching algorithm; screening and removing retained matching points, and keeping the coordinate of the matching point closest to the center of the license plate; performing stereoscopic measurement on the screened matching points to obtain a position under space coordinates; and calculating the travel distance of the target vehicle within a certain time interval to obtain the running speed of the target vehicle. Through adoption of the method, mounting and debugging are facilitated; a plurality of trained features can be automatically identified at the same time; and the development demand of a future intelligent traffic network and Internet of things can be met better.

Description

technical field [0001] The invention relates to the technical field of binocular stereo vision, in particular to an intelligent vehicle speed measurement method based on a binocular stereo vision system. Background technique [0002] Traditional traffic monitoring systems usually use camera equipment installed above or on the side of the road to take pictures of the road, and then cooperate with embedded coils or radar, laser radar, etc. to detect speeding behavior and take pictures to collect evidence, which requires the cooperation of multiple systems to complete The work of speed measurement, license plate recognition, evidence collection and recording, installation, laying and commissioning are relatively complicated. For other driving violations, such as answering the phone while driving, not wearing a seat belt, etc., it is usually necessary to manually check the video or image records for detection. In the face of more and more vehicles, the way of manually detecting...

AI Technical Summary

Problems solved by technology

[0004] In view of the technical problem that the existing vehicle speed measurement method needs to destroy the road and cannot measure all the vehicles in the field of vision, the present invention proposes a vehicle intelligent speed measurement method based on the binocular stereo vision system. It does not need to destroy the road buried coil and can measure all vehicle targets in the field of vision at the same time, and can complete vehicle speed measurement on undulating road sections and turning sections

Images