Obstacle recognition for ADAS using stereovision and snake models

Advanced Driver Assistance Systems (ADAS) is progressing into urban traffics. Existing ADAS is concentrated on obstacle detection, mainly for pedestrian or vehicle. Limited work has been conducted on multi-class obstacle classification. It has been recognized that object classification is essential for safety applications in urban traffic. This paper addresses this issue and aims to develop an approach for simultaneous detection and classification of multi-class obstacles. In the paper, stereovision is used to segment obstacles from traffic background by using distance measure, then active contour model is adopted to extract complete contour curve of the detected obstacles. Based on the contour extracted, object features including aspect ratio, area ratio and height are integrated for classifying object types including vehicles, pedestrian and other obstacles. The approach presented here was tested on substantial complex urban traffic images and the corresponding results prove the efficiency of the approach.