针对红外图像特点,提出了红外图像直方图双向均衡算法,该算法在进行直方图均衡化处理的同时,消除灰度冗余,并在整个显示范围内对图像灰度级进行等间距排列,从而获得灰度连续的红外图像,通过采用并行处理结构及流水线技术在FPGA内实现了该算法,解决了热像仪信号处理中输入大动态范围和显示输出的小动态范围的矛盾,实验表明该技术能够使红外图像的细节和清晰度得到明显的增强。 According to the characteristics of infrared images, a bidirectional equalization algorithm for infrared image histogram is proposed. This algorithm eliminates the gray-scale redundancy while performing histogram equalization and arranges the image grayscales at equal intervals within the whole display range. The grayscale continuous infrared image is obtained. The algorithm is implemented in FPGA through the parallel processing structure and pipeline technology, which solves the contradiction between the input dynamic range and the small dynamic range of the display output in the signal processing of the camera. Experiments show that the technology Can make the infrared image details and sharpness is significantly enhanced.