Vehicle-mounted grid map generation method and system based on fusion of laser radar and look-around camera

Abstract

The invention discloses a vehicle-mounted grid map generation method based on laser radar and look-around camera fusion, and a plurality of laser radars and look-around cameras are arranged to sense the surrounding environment of a vehicle. The method comprises the following steps: obtaining an image acquired by the look-around cameras at the moment t1, extracting a drivable area from the image, and carrying out binarization to obtain a binarized drivable area; projecting the binarized drivable area corresponding to each look-around camera to a vehicle coordinate system to obtain a vehicle aerial view; rasterizing a preset range around the aerial view vehicle to obtain MapCam and Maskcam, and converting the MapCam and Maskcam into a world coordinate system; obtaining laser point cloud datacollected by a laser radar at the moment t2, performing external parameter normalization on the data to a vehicle coordinate system, and obtaining point cloud data around the vehicle; rasterizing thepoint cloud data in the preset range, and removing ground point cloud to obtain obstacle point cloud; obtaining Maplidar and Masklidar based on the obstacle point cloud, and converting the Maplidar and the Masklidar into a world coordinate system; and obtaining a fused raster map Mapfusion according to the obtained raster map Mapfusion.

Description

technical field [0001] The invention relates to a method and system for generating a map, in particular to a method and system for generating a vehicle-mounted grid map. Background technique [0002] In recent years, with the rapid development of unmanned driving technology, the possibility of unmanned vehicles being applied in daily life is also increasing. In the field of unmanned vehicle technology, different application scenarios can be divided into low-speed and high-speed scenarios. High-speed scenarios are generally defined as high-speed or urban expressways. The types of traffic participants in the surrounding environment of autonomous vehicles are limited, generally limited to The target level information provided by vehicles or special markers such as cone barrels and environment perception sensors such as radar and cameras can meet the needs of environment perception. In low-speed scenes, due to the complex environment, it is impossible to predict what kind of ob...

AI Technical Summary

Problems solved by technology

[0005] Based on this, the present invention expects to obtain a vehicle-mounted grid map generation method and system based on the fusion of lidar and surround-view cameras, which can effectively improve the perception accuracy of autonomous vehicles for grid maps by fusing the surround-view camera and lidar , to solve the problem of blind spots in raster map generation

Images