Moving Target Detection And Tracking System Based On FPGA

Recently, with the continuous development of communications technology, computer technology, image processing techniques, computer vision has become one of the aspect of the current important subject. Moving target detection and tracking is an important part of computer vision research, combines many areas of pattern recognition, image processing, automatic control, and other advanced technology, which has important practical value in the area of industrial, military, medical, security monitoring.According to software-hardware partitioning method, this paper is based on FPGA as the core design of the moving target detection and tracking system to capture video images, moving target detection and tracking integrated on a smaller embedded systems. Compared with the DSP, ARM and other systems, FPGA has a wealth of internal resources, easy to upgrade the system and strong ability of parallel processing. This system takes EP2C70, Altera’s Cyclone II series as the core. Using an object detection algorithm, the designed system can find the moving objects in video streams which have been acquired through a CMOS image sensor controller. The video streams and detecting results by tracking algorithm to track moving target.This paper using a synergistic manner through software and hardware, large amount of computation by the hardware description language (Verilog HDL) to complete, such as video image capture, Bayer to RGB format conversion, multi-port SDRAM memory design, and image display control. So the presenting system has outstanding features such as real-time data processing. Software:the algorithms are programmed and debugged by C code in the Nios Ⅱ EDS environment to realize moving object detection and tracking algorithm. Algorithm is using frame substraction and centroid tracking algorithm for moving object detection and tracking. Using the system Nios II embedded technology is easy to extend and maintain this system, and under the code is easier to write algorithms using C language in Nios II EDS platform.This paper is tested many times that show this design has better anti-interference ability, real time and high accuracy on the indoor environment moving target detection and tracking.