基于模糊最大熵的红外图像增强算法

针对红外图像整体亮度偏暗、对比度较低、目标与背景区分不明显的特点,论文提出了一种基于模糊最大熵和改进的S函数的图像增强算法。该方法首先通过模糊最大熵准则求取最佳阈值,再使用改进的S函数对阈值两侧的像素灰度进行模糊域的非线性拉伸,从而得到增强后的红外图像。实验结果表明,该算法能明显提高红外图像对比度,突出目标并降低背景噪声。

Enhancement methods based on maximal fuzzy entropy for infrared images

Considering the characteristics of low luminance,low contrast and the inconspicuous difference between targets and backgrounds in infrared images,an enhancement method based on maximal fuzzy entropy and improved adjustment function is designed.To enhance the target area in infrared images,firstly,an optimal threshold is searched through the maximal Fuzzy-Entropy criterion,and a non-linear stretching method in fuzzy sets is carried out respectively to the pixels on both sides of the threshold.Results of the experiment indicate that the method could well improve the contrast of infrared images,the targets of which are outstanding from the backgrounds,thus strengthening the details of targets and suppressing background noises in the infrared images.