A Lidar Online Calibration Method for Autopilot System Fused with Vehicle Information

Abstract

The present invention discloses an on-line calibration method of laser radar for automatic driving system that integrates vehicle information, and its purpose is to provide an end-to-end online calibration method for external parameters of laser radar, avoiding complex mathematical model derivation and optimization, and avoiding the use of additional equipment , using the data analysis capability of the deep convolutional neural network to process the point cloud data of the laser radar and the vehicle ECU data, and estimate the external parameter error of the laser radar in real time online, so as to realize the real-time correction of the external parameters of the laser radar and improve the environment of the automatic driving system The accuracy and stability of the perception function ensures the driving safety of the automatic driving system.

Description

technical field [0001] The invention relates to a laser radar calibration method for an automatic driving system, in particular to an online laser radar calibration method for an automatic driving system that integrates vehicle information. Background technique [0002] The autonomous driving technology that has emerged in recent years is expected to solve the social problems of the high number of traffic accidents, and even change the way people travel and affect the design pattern of cities. Therefore, it has received attention from various countries and has become a research hotspot in the automotive industry. [0003] The autonomous driving system needs to interact with the environment, and thus needs to perceive the state of the environment. At present, the environment perception sensors commonly used in autonomous driving systems include: lidar, camera, millimeter wave radar and ultrasonic radar. Among them, lidar is favored because of its accurate distance measuremen...

AI Technical Summary

Problems solved by technology

[0007] Existing lidar external parameter calibration methods have the following defects: 1) Most of them are static calibration methods, which require special sites and special processes, and cannot be implemented online; 2) Data modeling and calculation are required, which increases the complexity of the algorithm and also The model mismatch error is introduced; 3) Iterative optimization is usually required to find the optimal solution, and it is easy to fall into local optimum, and the real accurate external parameters cannot be obtained

However, no related work has been carried out

Images