PEDESTRIAN DETECTION WITH RADAR AND COMPUTER VISION

Abstract

This paper presents a method for detecting pedestrians on-board a moving vehicle. The perception of the environment is performed through the fusion of an automotive radar sensor and a monocular vision system. The fusion uses a two-step approach for efficient object detection. In the first step, a target-list is generated from the radar sensor. The items in the list are hypotheses for the presents of pedestrians. In a second step, hypotheses are proved by the vision system. This method achieves very large speed-ups compared to a sole image processing solution.