The Design Of Image Recognition And Tracking System Based On FPGA

In image processing, image recognition and tracking are difficult and critical technologies. This technology is widely applied to military, industrial and other fields. Image recognition and tracking algorithms can be implemented in software or hardware. Out of cost considerations, many image processing system be implemented witch software methods, but in high-speed real-time systems and the processing of high-resolution systems, using hardware methods to process image is an inevitable choice. With the development of embedded technology, image recognition and tracking technology to become a low-cost, low power, small size system. So this paper, we use the hardware method to design the Image Recognition and Tracking System.There are three ways to achieve the system witch hardware method: the integration of specialized image processing chip, DSP digital signal processor or FPGA with the relevant circuit. Compare these methods, FPGA chip has rich hardware resources and hardware reconfiguration. The recognition and tracking system can greatly improve the flexibility and versatility, when use the FPGA chip.This paper focuses on intelligent tracking system, and research the system hardware and software algorithms of the target object identification and tracking system. We build a image recognition and tracking system based on FPGA, full use the high-speed of FPGA parallel processing and ease development process. And FPGA has online compiler debugging, so that the image recognition and tracking system can reduce the cost, shorten the development cycle and be able to meet the real-time requirements. Combination of hardware circuit design technical, we design the hardware circuit of image tracking system based on FPGA, and we research the key technologies in-depth. This paper analyzes the design flow of FPGA , complete the control module software. The system realized the module of image capture, color information conversion, image information storage and display. We use the Sobel edge detection algorithm in the image processing, and combined with color information to achieve the tracking and display real-time. Experimental results show that this system complete the target object identification and tracking...