基于区域分割的自适应MSR方法

Adaptive MSR Method Based on Region Segmentation

提出一种基于区域分割的自适应MSR图像增强方法,通过将图像分割为不同光照区域,对明暗区域边界的光照进行自适应估计,使用区域信息修改滤波器系数来降低光照在不同区域间的扩散,从而达到抑制光晕的目的,并采用改进的MSR方法提升图像的视觉质量。实验结果和数据分析证明了该方法是有效的。

This paper presents an adaptive Multi Scale Retinex（MSR） method based on region segmentation.The method of image segmentation by region for the different illumination of the scene,which has illumination changes spatial acutely on the edge.The modified filter takes the regional information into account to reduce the coefficient illumination between the different regions,so as to achieve the purpose of inhibition halo.The improved MSR method achieves visual quality enhanced images as well.Through the experimental results and data analysis,this method is approved to be effective,and provides satisfied result.