A vehicle longitudinal positioning system and method based on variable grid image feature detection in automatic driving

Abstract

The present invention provides a vehicle longitudinal positioning method based on variable grid image feature detection in automatic driving. According to the target ahead obtained from the data output by the high-precision navigation system and the target distance calculated by the road in the high-precision navigation, through application The ORB feature extraction algorithm based on the variable grid area (carrying scale information) can retrieve specific forward targets in the vehicle-mounted binocular vision system and output the distance from the visual system to the forward target. Based on this distance and the installation position of the binocular system in the vehicle, it can be corrected with the vehicle trajectory in high-precision navigation to improve the longitudinal positioning accuracy of the vehicle in autonomous driving.

Description

technical field [0001] The invention belongs to the technical field of automobile automatic driving, and relates to a system and method for self-positioning of an automatic driving vehicle, in particular to a system and method for longitudinal positioning of a vehicle in automatic driving based on variable grid image feature detection . Background technique [0002] The automatic driving system of intelligent vehicles needs to rely on high-precision map data. According to the specified destination information, it can dynamically and economically complete global and local path planning to form its own navigation trajectory, and complete various control actions of unmanned vehicles safely and conveniently. In the implementation process of this system, it is necessary to know the high-precision positioning information of the vehicle itself in real time and accurately, so as to make a decision-making judgment on the current driving state control. [0003] The conventional posit...

AI Technical Summary

Problems solved by technology

However, the visual odometer based on the binocular camera needs to perform real-time correction, registration, and calculation of the disparity map of the left and right images, and cannot output at a high frequency. For images with a size of 1600×1200 pixels, the depth calculation frame of the binocular image is performed. The rate is less than 10FPS(Frames Per Second)

The frame rate of image processing is low, and the output frequency of positioning information calculated by the camera is also low. When it is fused with other positioning information such as GNSS+IMU, more issues such as time synchronization and linear / nonlinear interpolation need to be considered. Affect the reliability, real-time, and accuracy of high-precision positioning information

Images