Vehicle detection based on distributed sensor decision fusion

Obstacle detection is one of the key functions to intelligent vehicle (IV). In this paper, the recent work of the IV research group of Jilin University of China in detecting the preceding vehicle is introduced. First, a brief overview is given of the vehicle detection approach based on a CCD camera and laser radar, respectively. Then a new vehicle detection method based on distributed bi-sensor decision fusion is put forward. Based on the minimum Bayesian risk criterion and fuzzy a priori probability and fuzzy cost function, the binary decision fusion rule is established. Experiment results are also presented.