摘要： 针对彩色图像进行光照补偿后难以保持色彩恒常性的问题，提出一种基于Retinex和低通(LP)滤波的彩色图像光照补偿方法。将图像从RGB空间转换到HSV空间，保持H和S分量不变，对V分量使用LP滤波器V分量的高频成分H和低频成分VL，对VL采用 Retinex算法进行光照补偿，利用VH和经光照补偿的VL及双重框架运算的LP重构算法进行重建得到V分量，将图像反变换至RGB空间得到最终的增强结果。实验结果表明，该光照补偿方法能对图像进行有效的光照补偿，保持较好的颜色恒常性。

关键词: 色彩恒常性, 光照补偿, 视网膜皮质模型, 低通滤波, 双重框架

Abstract: To solve the conundrum that it is difficult to preserve the color constancy after doing illumination compensation, a novel algorithm based on Low-Pass(LP) filter and Retinex is proposed. In this algorithm, RGB image to be dealt is transformed into HSV space at the very beginning, the H and the S element are kept steady, then by using the LP filter the illumination image V is divided into high frequency and low frequency parts, which are symbolized by VH and VL respectively, in the flowing steps Retinex algorithm is used on the VL image to do illumination compensation, VH and the compensated VL are used to reconstructed V, and the compensated image in HSV space is transformed back to RGB space, where it gets the final image. Through experiment, it is confirmed that the proposed algorithm to do color image’s illumination compensation yields some good results both on illumination compensation and on preserving the color constancy.

Key words: color constancy, illumination compensation, Retinex model, Low-Pass(LP) filtering, bi-framework

中图分类号: 

• TP391