多维直方图联合概率分布的掩模矩阵降维算法

Algorithm with Mask Matrix to Reduce Dimension for Multi-dimension Histogram Jointed Probability Distribution

彩色图像的分量独立处理或感知彩色空间的强度分量处理,都能达到增强图像对比度的效果,但由于没有考虑各个分量之间的联系,不能开发完全的RGB色域。多通道直方图建模技术可以改善上述缺陷,就是通过求取多个分量的直方图联合概率密度函数和累积分布函数,从而得到彩色图像的像素灰度值映射函数,但计算联合直方图的过程也导致了过高的运算复杂度。为了降低计算量,提出了掩模降维的方法,将计算多维直方图的过程转化为了计算一维直方图,从而大大降低了运算复杂度,提高运算速度,增强对比度的效果。

Component - wise processing and intensity processing of perceptive color space for color image all can enhance image contrast. However, neither inter - dependency of components is considered, nor is full RGB gamut exploited completely by two methods. Multi - channel modeling can improve this defection. Multi - channel modeling calculates the histogram associated with probability density function and cumulative distribution function, and then obtains the mapping function of pixel value of color image. But, jointed channel processing leads to huge operation complexity. This paper proposes reduced dimension with mask matrix, consequently multi - dimension histogram processing is converted to calculate one dimension histogram. Finally computation complexity is remarkably decreased.