Research Of Algorithm For Infrared Image Processing Based On FPGA

Along with the infrared technology is developing rapidly, it has more and more widely used in military, medicine, monitor, and People's Daily life. Because of the infrared imaging environmental complexity and detector pieces, limitation of the infrared imaging system imaging effects, insufficient in civil detection and other applications, the main performance for target and background contrast is bad, the image edge blur, noise larger shortcomings. For infrared image is more suitable for human eye observation, suitable for later target identification and tracking process, it is necessary to further treatment to image enhancement.With the fast development of the computer technology and integrated circuit technology, not only the algorithm and system instruct of image processing area improve fast. Meanwhile, the developments of large-scale programmable logic device provide a hardware-based for the real time image processing. Infrared image enhancement was realized based on FPGA platform in this project.Based on the analyses of several algorithms to digital image enhancement and the characteristic of infrared image, two infrared image enhancement algorithms fit for hardware design are presented. They are adaptive platform histogram equalization algorithm and quick median filtering algorithm, the algorithm through MATLAB simulation validation, and realized at FPGA image processing platform .Then based on the characteristic of structure of FPGA/ASIC, designing scheme of infrared image processing is put forward which has a Altera EP2C8 FPGA, the digital CMOS image sensors with MICRON company, two SDRAM chips as a cache and USB2.0 interface CY7C68013 chip. Design and implement of transmit module between FPGA and SDRAM, control module between FPGA and USB2.0 interface CY7C68013 and bus switch module are done by using QUARTUSII8.1. The implement of USB2.0 interface CY7C68013 firmware programmed is done on KEILC. The adaptive plateau histogram equalization algorithm and quick median filtering algorithm of infrared image are realized on FPGA platform using Verilog HDL. IP micro function module. They are complied tested on the FPGA.Aimed at the phenomenon of gray-level truncation and losing details, a new method of infrared image histogram equalization with image details preservation is proposed. It is the key to choose the threshold in the plateau histogram equalization, through choosing of image histograms mean lower threshold as to different platform of infrared image can adaptive adjustments platform threshold, this algorithm can highlight weak targets and details, avoid gray faults and details are lost. Some algorithms about infrared image processing are studied, then comparing their advantage and disadvantage, it is provided that fast median filter on FPGA for image low-level processing and its hardware structure diagram, then realizes it by Verilog HDL language. It results as this fast median filter can lower the system resources share and promote system.At last, the work is summarized and prospected.