为提高某中波红外探测器的图像质量, 设计了基于FPGA的红外图像实时处理系统, 系统能够完成实时的非均匀性校正与盲元补偿处理。介绍了目前常用的非均匀性校正、盲元识别和补偿算法, 并结合实际工程需求采用多点法进行非均匀性校正以及8点平均法进行盲元补偿。在仿真实验成功的基础上, 基于FPGA平台构建了硬件平台。系统可以实现系数自定义更新, 可以手动或自动完成非均匀性较正系数计算, 以及实现盲元列表的自动更新操作。利用某国产中波红外探测器对处理系统进行了测试试验, 实验结果表明: 校正后图像非均匀性＜03%, 盲元率＜0001%。系统工作稳定、可靠, 图像处理满足实时性和精度要求。

In order to improve the image quality of certain medium wave infrared detector, a infrared image real-time processing system based on FPGA is designed, which can accomplish nonuniformity correction and blind pixel compensation in real time. Some widely used nonuniformity correction algorithms, blind pixel detection and compensation algorithms are introduced. Combined with actual project requirements, nonuniformity correction with multi-point method and blind pixel compensation with eight-point average method are realized on the FPGA platform on the basis of successful simulation. User-defined coefficient updating, non-correction coefficients manually or automatically calculating and blind pixel lists automatically updating are realized by the system. The processing system is tested using certain domestic medium wave infrared detector, and the result shows that the nonuniformity of corrected image is less than 03% and the blind pixel of corrected image is less than 0001%. The system is stable and reliable, which satisfies the requirements of image processing in real-time and accurancy.