Lane validity prediction method and prediction system based on multi-dimensional features

Abstract

The invention provides a lane effectiveness prediction method and prediction system based on multi-dimensional features. The method comprises steps of supplementing and adjusting a fragmented perception lane marking line based on a crowdsourcing trajectory, and obtaining a smooth and complete lane sideline; constructing lanes by using adjacent left and right lane marking lines, and dividing road intervals at positions where the number of the lanes changes along the track driving direction to obtain a plurality of road intervals; counting five-dimensional feature vectors related to each road interval; and constructing a lane validity prediction model based on the five-dimensional feature vector, and predicting lane validity. According to information such as distribution density of tracks in a lane, spatial distribution of the lane on a road surface, perception proportion and supplementary proportion of left and right lane sidelines and the like, modeling is carried out through the basic characteristics, and credibility of existence of the lane is predicted, so invalid lanes generated due to offset tracks or sparse tracks are removed; and precision and the safety of a high-precision map are improved.

Description

technical field [0001] The present invention relates to the field of high-precision maps, and more specifically, to a lane effectiveness prediction method and prediction system based on multi-dimensional features. Background technique [0002] Building a crowdsourced map by collecting data from batches of low-cost crowdsourcing collection vehicles is one of the mainstream methods for high-precision map formation. Lines are supplemented and adjusted to build a lane-level high-precision map. [0003] Since the collection equipment of the crowdsourcing collection vehicle itself is low-cost, it is relatively normal to have trajectories drifting in environments such as building shelters and bad weather. For abnormal trajectories, vehicle kinematics and Kalman filtering can be combined. Filtering is performed, but for those trajectories that deviate slightly from the lane, the filtering effect is not so ideal. The lane sidelines speculated by these deviated trajectories and perc...

AI Technical Summary

Problems solved by technology

The above methods rely solely on the trajectory density and ignore the connectivity of the road network. In fact, the topological connectivity of the lanes reflected by the sparse trajectory may be crucial.

Images