针对红外图像灰度分布集中、对比度低的特征, 提出了一种基于改进直方图均衡的对比度增强算法。首先采用线性对比度增强将原始16位红外图像映射到8位图像A；然后采用改进的平台直方图均衡将原始16位红外图像映射到8位图像B；再根据输入图像的灰度级范围动态确定映射图像A和B的权值；最后以确定的权值将映射图像A和B合并, 得到最终对比度增强的图像。该方法克服了传统平台直方图均衡算法噪声过大及亮度突变的缺点, 动态结合了传统的灰度变换增强算法, 能根据全图目标与背景灰度的分布情况自适应调整对比度。实验表明, 该算法在增强目标对比度的同时有效保留了图像的整体信息, 改善了视觉效果。

Aiming at the features of concentrated gray-level distribution and low contrast of infrared images, a contrast enhancement algorithm based on an improved histogram equalization method is proposed. Firstly, a raw 16 bit image is mapped to 8 bit image A through linear contrast enhancement. Secondly, the raw 16 bit image is mapped to 8 bit image B by using the improved platform histogram equalization method. Thirdly, the weights of image A and image B are determined according to the gray-level range of the input images. Finally, the determined weights are used to combine image A and image B into the final contrast-enhanced image. This method has overcome the drawbacks of great noise and light mutation in traditional platform histogramequalization algorithms. Because the traditional gray-level transform enhancement algorithm is incorporated, it can adjust the contrast of the image adaptively according to the gray-level distribution of the object and background in the whole image. The experimental result shows that the proposed algorithm can keep the entire information in the image effectively while enhancing the contrast of the object.