A road boundary detection and tracking method based on a three-dimensional laser radar

Abstract

The invention relates to a road boundary detection and tracking method based on a three-dimensional laser radar. The method comprises the following steps: converting polar coordinate data of a laser radar scanning point into rectangular coordinate data; Extracting ground points by using a random sampling consistency plane fitting method and a principal component analysis method; Roughly extractingroad boundary feature points by adopting a multi-feature, wide-threshold and hierarchical feature point extraction method; Dividing the roughly extracted road boundary feature points into a left roadboundary feature point and a right road boundary feature point by using a projection-based method by referring to the idea of linear discriminant analysis; Respectively filtering the left road boundary point and the right road boundary point by using a random sampling consistency method to obtain final road boundary points; Fitting the left road boundary line and the right road boundary line by adopting a quadratic polynomial random sampling consistency method; And tracking the road boundary line by adopting a Kalman filtering method based on amplitude filtering. The method can be used for processing various complex scenes and working conditions, and can be widely applied to unmanned driving.

Description

technical field [0001] The invention belongs to the technical field of unmanned vehicle environment perception, and specifically relates to a road boundary detection and tracking method based on a three-dimensional laser radar. Background technique [0002] Since the birth of autonomous driving, environmental perception technology based on multiple sensors has attracted the attention of many scholars and research institutions. Among them, road boundary detection is an important research content in unmanned driving environment perception. The road boundary can separate the road area from the non-road area, thus providing an important basis for the decision-making and planning of unmanned vehicles; through the road boundary, the invalid sensor data within the range of the road boundary can be filtered out, and the efficiency of subsequent obstacle detection can be improved; because The obstruction of GPS signals by trees, overpasses and tunnels in urban environments, and road...

AI Technical Summary

Problems solved by technology

[0004] The camera is small in size and low in price and can provide rich texture and color information, but it is easily affected by light and weather, and the monocular camera has no depth information

Although the stereo camera has depth information, the field of view is small and the parallax needs to be calculated, so the real-time performance is poor

Although 2D lidar is not affected by changes in external lighting, it can directly obtain depth information, but it can only scan at a fixed angle, the amount of data is small, and it is easily affected by observation noise

[0005] Although 3D LiDAR acquires data quickly and with high precision, the amount of data is huge and disorderly, which puts high demands on the algorithms for processing 3D LiDAR data

This method is complex and it is easy to filter out effective boundary feature points when the boundary feature points are discontinuous or the road is curved, and it is easily interfered by obstacles in the road

Kunlei Hu et al. extract feature points through curvature features and height difference features, and then directly fit the road boundary through RANSAC straight line. This method is easily affected by obstacles with the same characteristics as the road boundary inside and outside the road, such as surrounding roads in the same direction as the road. railroad tracks would be seriously disturbed

Images