METHOD AND DEVICE FOR RECOGNIZING ROAD SCENE AS WELL AS RELEVANT PROGRAM AND NON-TRANSITORY COMPUTER-READABLE MEDIUM

摘要

Disclosed is a road line recognition method comprising steps of dividing a disparity map including a road scene into plural sub image regions (S210); extracting, regarding each sub image region, a feature vector generated based on projection information of respective pixels therein on each plane defined by a world coordinate system (S220); dividing the disparity map into plural cells having a predetermined size so as to determine sub feature vectors corresponding to the plural cells according to the feature vectors corresponding to the plural sub image regions (S230); and recognizing, by utilizing a road scene classifier pre-trained according to a statistical learning approach, the road scene based on the sub feature vectors corresponding to the plural cells (S240).