A new vehicle detection algorithm for real-time image processing system

Abstract

Target detection system based on visible image sensor is applied in many fields. In vehicle detection field, people are now paying more and more attention to the research on finding a real-time vehicle detection system and algorithm with preferable precision. This paper designs a real-time system based on digital signal processing (DSP) TMS320DM643 and proposes a fast and robust corner detection algorithm for vehicle detection. The new corner detector is based on a novel mask which uses only the margin information of two circles. Then the features of corners are extracts to recognize vehicle target. The experiments show that the developed DSP system and the vehicle detection algorithm can fulfill the requirement of real-time applications and can be effective to achieve a higher precision for vehicle detection.