弱光夜视是短波红外成像的重要应用领域之一。针对短波红外弱光图像对比度低, 增强后噪声也被放大的特点, 提出了一种基于小波变换与像元对目标的短波红外图像增强算法。首先通过小波变换获得不同频率成分的子带图像; 然后对低频子带图像进行基于像元对目标的灰度变换处理, 对高频子带图像进行可变阈值降噪处理; 最后通过小波反变换将处理后的子带重构得到增强结果。将该算法与基于直方图的增强算法, 全局优化线性窗口色调映射算法和自然保持增强算法进行比较, 采用图像的信息熵和基于Michelson法则的对比度增强度量作为客观评价指标, 结果表明本文算法更为有效地提高了短波红外弱光图像的对比度, 抑制了噪声的增强, 提升了图像的视觉效果。

Low light night vision is one of the important application areas of SWIR imaging. Aiming at low contrast of SWIR night images and the characteristic that noise will be also amplified after enhancement, a SWIR low light image enhancement algorithm is proposed based on wavelet transform and pairwise target contrast. Firstly, the sub-band images of different frequency components are obtained by wavelet transform; then the low-frequency sub-band images are transformed based on pairwise target contrast, and the high-frequency sub-band images are subjected to variable threshold noise reduction processing. The processed sub-band images are reconstructed by inverse wavelet transform to obtain an enhanced image. The proposed algorithm is compared with the enhancement algorithm based on histogram, globally optimized linear windowed tone mapping algorithm, and naturalness preserved enhancement algorithm. The information entropy and measure of enhancement are used as objective evaluation indicators. The results show that the proposed algorithm can effectively improve the contrast, suppress noise amplification and improve the visual effect of SWIR low light image.