Abstract: With constraints to the performance of the IR detector，IR image usually has lower visual effect with low contrast and weak detail information.It needs to be enhanced to improve the quality.A detail enhancement method for infrared image is proposed in this paper.Firstly，the extremum points of the original image are detected and replaced to eliminate the shot noise.Secondly，the detail information is achieved after the low-passed image is subtracted from the image acquired in the first step.Meanwhile，the gradient information of the image is obtained with morphological approach.Together with the original grey value，a weight value for grey-scale statistic is figured out.Finally，the histogram equalization algorithm based on the weight value is adopted to get high-quality image with distinct details.Experimental results demonstrate that the method has satisfactory performance with low computation complexity.

Key words: infrared image, weight value, histogram equalization, detail enhancement

摘要： 受焦平面阵列（FPA）性能限制，红外图像对比度低，细节信息不明显，视觉效果较差，需要经过增强处理改善图像质量。提出了一种红外图像细节增强算法，首先检测原始图像中的极值点并进行替换，去除散粒噪声，再用消噪后图像与低通滤波图像相减得到细节信息，同时还通过数学形态学方法得到图像梯度信息，结合原灰度值计算出灰度统计权值，最后将此权值应用于细节图像的直方图均衡算法中，得到细节增强后图像。实验表明此算法具有很好的效果，而且计算速度快。

关键词: 红外图像, 权值, 直方图均衡, 细节增强