动态范围压缩和细节增强是红外图像处理的两个重要课题。为了将高动态海面背景红外图像清晰显示, 提出一种高动态范围压缩及细节增强算法。首先, 通过基于梯度边缘信息的多方向拉普拉斯增强方法, 将梯度图像平滑处理, 并与多方向拉普拉斯滤波相乘, 实现高动态范围图像的细节增强; 然后统计增强后图像的动态广义直方图信息; 最后采用灰度级分组的方法构造映射函数, 将高动态范围压缩到8 bits, 输出可清晰显示的红外图像。对大量海面背景红外图像进行实验分析, 结果表明, 该算法提高了图像的对比度, 有效增强了舰船目标细节, 同时抑制了海面背景噪声的放大和光晕现象的产生, 最终获得较好的输出图像。

Dynamic range compression and detail enhancement are two important issues for effectively displaying high dynamic range infrared(IR) images on standard dynamic range monitors. Sophisticated techniques are required in order to improve the visibility of the details without introducing distortions. To clearly show the highly dynamic infrared image of sea background, a high dynamic range compression and detail enhancement algorithm was presented which was made of three main steps. First, highly dynamic infrared image was enhanced by a multi-directional Laplacian enhancement method based on gradient edge information, and the first order gradient image smoothed was combined with multi-directional Laplacian filtering to enhance images. Second, the dynamic generalized histogram information of the enhanced image was obtained. Finally, the gray level grouping method was used to construct the mapping function, which maps 14 bits intensity in the input image to 8 bits intensity in the output image. As a result, infrared image with low dynamic range was obtained clearly. Experiments on a large number of sea-surface infrared images were conducted. The results verify that the proposed algorithm can improve the contrast of the image, effectively enhance the ship target details, suppress background noise amplification, and avoid the generation of halos. Therefore, the image with high quality was achieved.