一种基于双滤波的空间图像色域映射算法

摘要

针对逐点色域映射方法在图像色域映射时出现的细节损失较大的问题，本文提出了一种新的空间图像色域映射方法。新方法采用双滤波技术将输入图像分解为对应边缘轮廓信息的基础层图像和对应纹理细节信息的细节层图像，先对基础层图像进行彩度优先的逐点色域裁剪，然后将细节层信息补偿给色域裁剪后的图像。最后对细节补偿后的图像进行亮度优先的色域裁剪，从而得到最终的映射图像。本文还分析了双滤波参数设置对色域映射结果和光晕的影响，得出了合理的双滤波参数设置。通过心理物理学实验可以验证：新方法的性能与经典的细节补偿类映射方法相当。另外，新方法在色域映射过程中还能更加有效地抑制光晕的产生。%A spatial Gamut Mapping Algorithm(GMA)was proposed in order to avoid the detail loss produced by pointwise GMAs.Firstly,the bilateral filter was employed to decompose the original image into a base layer image(contai-ning image-edge information)and a detail layer image(containing image-texture information).Then the base layer image was gamut-clipped by the pointwise GMA that emphasizes chroma over luminance preservation.After the first gamut map-ping process,the detail layer was added to the gamut-clipped base layer image.Finally,the merged image was gamut-clipped again by another pointwise GMA that emphasizes luminance over chroma preservation to get the final gamut-mapped image. The influences of different bilateral filter parameters on gamut mapping results and halo-artifacts were analyzed.The psycho-physical experimental results show that the proposed spatial GMA with proper bilateral filtering parameters performs signifi-cantly better than standard pointwise GMAs and has almost same performance as another classical spatial GMA.In addition, the proposed spatial GMA can reduce the halo-artifacts much more effectively.