Multi-Sensor Fusion based Target Detection using EO/ SAR

Abstract:
Multi-sensor data fusion can increase the target detection probability in the rapidly changing environment of the UAV (Unmanned Aerial Vehicle) surveillance system. In this paper, Decision level sensor fusion technique using both EO (Electro Optical) and SAR (Synthetic Aperture Radar) images is proposed for improving the target detection probability compared to the case of the conventional single sensor. The proposed fusion algorithm requires the same sets of EO and SAR image, separately taken from the same scene at the same time. After sensor level target detection, the decision level fusion is performed for the combined target detection. As a result, the number of false alarms per imaging area (km²) is significantly reduced down to 14~23% compared to the case of an individual sensor, while the probability of target detection is well maintained up to 85%. This technique can be effectively used for the application of the UAV surveillance systems.