Study on shipborne video electro-optical tracking system based on FPGA

Abstract

Shipborne video electro-optical tracking system can be used for ocean target detection and recognition, marine rescue, etc. In this paper, an FPGA is selected as the core processor to realize the target detection and servo tracking. We propose a colour matching centroid tracking algorithm to solve the impact of complex water surface changes. Test results show that the system works normally and stably. In the 640 × 480 video image resolution, the tracking accuracy of the azimuth motor is 80 pixels, and the maximum tracking angular velocity is 30°/s. The tracking accuracy of the pitch motor is 50 pixels, and the maximum tracking angular velocity is 10°/s. The minimum size of target recognition is 16 × 16 pixels matrix.