基于积分图像的灰度分布标准化快速算法  

Rapid Algorithm of Grayscale Distribution Normalization Based on Integral Image

作　　者：邵平[1] 杨路明[1] 黄海滨[1] 

机构地区：[1]中南大学信息科学与工程学院

出　　处：《计算机应用研究》2007年第3期277-279,共3页Application Research of Computers

基　　金：广西教育厅资助项目(桂教科研[2004]20号);玉林师范学院重点科研项目(05DWX16);玉林师范学院科研项目(05YBWX23)

摘　　要：在积分图像基础上提出和应用平方积分图像,实现了一种计算速度与图像窗口大小无关的快速灰度均值和方差算法。应用该算法,结合对灰度分布标准化近似计算公式的变换,实现了图像窗口灰度分布标准化的快速计算。实验证明,当需要对一幅图像的大量图像窗口进行灰度分布标准化时,这种算法对提高计算速度十分有效。Based on integral image, the square integral image was proposed and used to realize a rapid algorithm for calculating grayscale average and variance of image windows, and the calculating speed was independent of the windows ＇ size. With such an algorithm and the transform of approximate formula for calculating grayscale distribution normalization, a rapid algorithm for grayscale distribution normalization of image windows was realized. It was proved to be very good for improving the computing speed when a large number of image windows have to be processed with grayscale distribution normalization.

关 键 词：积分图像 灰度分布标准化 方差 快速算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]