为实现室外模糊红外图像对比度增强，提出一种基于去雾模型的红外图像增强方法。结合红外图像特点，对可见光去雾增强方法进行改进优化。采用三级高斯金字塔分解扩展实现图像快速均值滤波，获取透射率粗估计；通过图像统计信息自适应细化透射率，恢复出无雾图像；针对无雾图像整体亮度较暗现象，进一步采用背景抑制的分段对比度增强。实验结果表明：增强后图像细节信息突出，层次感丰富，人眼视觉效果良好。客观测评结果表明，该算法能有效增强红外图像对比度。嵌入式平台测试耗时28ms，可以实现实时红外图像增强处理。

In order to enhance the outdoor blurred infrared image contrast, an infrared image enhancement based on haze remove method is proposed. The novel algorithm optimizes and improves the visual image haze remove method which combines the characteristics of the infrared images. In order to get the transmission rate coarse estimation, a fast average filtering using three levels Gaussian pyramid operation is presented. The haze free image is recovered through self- adaptive transmission rate calculated with the statistics information of image. To deal with low luminance problem of the whole haze free image, a sectional contrast enhancement way is proposed which is capable of background suppression. Experimental results show that the enhanced infrared image has more detail information and stronger gradient than the original and has a perfect visual effects. The objective evaluation parameters illustrate that the contrast of haze free infrared image increases effectively by the proposed algorithm. The proposed algorithm can realize real time infrared image enhancement processing, as embedded platform test takes 28 ms.