Abstract:

In order to solve the problem of the loss of details in the low-exposure area and the decrease of visual perception due to the poor color distortion and saturation in the mixed exposure imaging algorithm, a Multi-Scale Retinex with Color Restoration（MSRCR） is proposed based on the multi-scale weight assessment.Firstly, based on the Retinex theory model, the image to be merged is decomposed into brightness component and reflected light component, and the normalized function of illumination compensation is constructed by combining the brightness component and ACES function for processing. The color recovery function is added to the reflected light component to improve the color details.Secondly, the image fusion weight values are designed from the four scales of exposure, saturation, contrast and color gamut, respectively, and the fusion ratio is optimized through multi-scale evaluation.Finally, the Laplacian pyramid fusion algorithm is used for multi-scale weight fusion to obtain the final image.Compared with the traditional image fusion algorithm, the experimental results show that this algorithm has better processing effect, effectively reduces the shadow distortion rate and improves the fidelity of visual information.

Key words: high dynamic range, Multi-Scale Retinex with Color Restoration（MSRCR）, lighting compensation, multi-scale assessment, pyramid fusion

摘要：

针对混合曝光成像算法过程中会出现低曝光处细节丢失且颜色失真饱和度不佳导致视觉观感下降的问题，提出一种多尺度权重评估的MSRCR（Multi-Scale Retinex with Color Restoration，MSRCR）混合曝光融合算法。基于Retinex模型将待融合图像分解为亮度分量与反射光分量，对亮度分量结合ACES函数构造光照补偿归一化函数进行处理，对反射光分量加入颜色恢复函数提升色彩细节；分别从曝光量、饱和度、对比度、色域四个尺度设计图像融合权重值，通过多尺度评估优化融合比例；利用Laplacian金字塔融合算法进行多尺度权重融合获得最终图像。实验结果表明，与传统的图像融合算法相比，该算法处理效果较好，有效降低了暗处失真率，提升了视觉信息保真度。

关键词: 高动态范围, 带色彩恢复的多尺度视网膜增强算法（MSRCR）, 光照补偿, 多尺度评估, 金字塔融合