基于高斯分布的自适应红外图像增强算法

An Adaptive Enhancement Algorithm Based on Gaussian Distribution for Infrared Image

针对红外图像信噪比和对比度低,结合红外图像直方图分布的特点,提出了一种基于高斯分布的自适应红外图像增强算法。该算法采用高斯分布函数平滑红外图像直方图中的主峰,以抑制红外图像背景和噪声的加强。算法中的高斯分布函数可根据原红外图像直方图的分布情况自适应产生,不需要人为设定,将所生成的高斯分布函数对原直方图进行转换后再进行直方图均衡处理,即可实现红外图像的增强。实验结果表明,该算法能有效地突显红外图像的目标,抑制其背景,具有较好的增强效果。

As for the low signal to noise ratio （SNR） and contrast of infrared image,an adaptive enhancement algorithm based on Gaussian distribution for infrared image is presented by the characteristic of infrared image histogram,in which the main peak of infrared image histogram was smoothed by Gaussian distribution function to inhibit enhancement of the background and noise.The Gaussian distribution function could be adaptively produced by the distribution of original infrared image histogram,no need for any parameters.Then,the conventional histogram equalization was performed on the original infrared image whose histogram was converted by the Gaussian distribution function.The experiment results show that the enhancement algorithm can not only enhance the objects effectively,but also constrain the background and noise greatly in infrared images.