Research And Implementation Of Target Recognition Technology Based On FPGA

With the development of science and technology, the target recognition technology isbecoming more and more important. The target recognition algorithm directly determines thefollow-up design’s accuracyand efficiency. As a part of image processing technology, imageacquisition and processing of target is developing in the manner of high speed, high integration,high precision and real-time. It has been applied in the fields of industrial, military, medical,and even families, such as industrial control, pattern recognition, surveillance systems, and soon.How to design a low-cost, data capacity, reliable, real-time target recognition system hasbecome a hot topic in image processing research field. Nowadays, the rapid development ofFPGA provides a new solution for it. This paper puts forward a kind of FPGA based on thetarget recognition scheme which is at the stage of current image processing technology. Thisscheme make the whole system divide in seven modules: image acquisition, imagepre-processing, image storage, data exchange, color recognition, edge detection and targetstatistics. The divisions of labor after the image acquisition are as follows:1. Image pre-processing including design of the signal control module in decoder chip andcolor space conversion module.2. Design of the Wishbone bus and SRAM controlled by FPGA.3. Design of color recognition and edge detection algorithms.4. Design of target statistics module.The effect of the image processing is tested in the experiment using a small ball in thesystem eventually. The study has laid a solid foundation for the further development of thefollow-up soccer robot while it can also be directly applied in the daily monitoring andcontrol system and it has a certain reference value for high-speed image processing andtesting on other occasions.