Research And Implementation Of Fast Image Processing Algorithms Based On FPGA

Digital image processing is closely related to the development of computer, multimedia, intelligent robot, expert system and other technologies. Today, along with the rapid development trend of computer intelligent technology in digital image processing,the digital image processing technology has been tied to a large extent to various fields, such as used in life science, aerospace, satellite remote sensing, industrial, military and public security and so on. Therefore, the real-time and efficient digital image processing system design is becoming increasingly important, so, the traditional design way of digital image processing technology by software has been unable to meet the designer’s needs very well. On the other hand, with the rapid development of semiconductor technology, silicon processing technology has entered the deep submicron field less than 30 nm, and multi-million gate circuits or over ten million gate circuits can be integrated on a single chip, therefore, the speed problem of digital image processing technology can be solved well by using various hardware development tools. Combining with the parallel execution and fast advantages of FPGA, this paper completed the research and implementation of digital image processing algorithms to achieve the purpose of rapid digital image processing on the FPGA hardware development platform.In order to realize this project to optimize image processing algorithms(the typical spatial filters and edge detection algorithms) in time, and the algorithms are implemented on FPGA by using Verilog HDL, and a design of FPGA- based parallel twin template theory to shorten image processing time has been completed based on the high-speed hardware computing power and the advantage of parallel computing capacity of FPGA, so as to achieve the purpose of rapid image processing FPGA-based.In order to verify the thought that the FPGA-based parallel twin template theory can improve the time performance of digital image processing algorithms, and design a number of comparative experiments, Among them, spatial filtering algorithms use the Lena image as experimental subject, and edge detection algorithms use bridge crack image as experimental subject, compare the experimental results with the FPGA-based parallel twin template theory, the FPGA-based single template theory and software(matlab). According to the characteristics of various image processing algorithms, acquiring the objective evaluation index values of image processing results by the experiments, and analyzing these results. Finally, under the condition of ensuring the precision of image processing algorithms, we verified the FPGA-based parallel twin template theory can improve the time performance of digital image processing algorithms, and made the best of the parallel computing capacity of FPGA.