摘要
提出了一种结合高维空间几何计算的图像消影方法,通过一组光滑滤波核与原始图像作用,生成一幅新的特征图像点,将此2个图像映射成为高维空间中的2个点,测算空间中点与特征图像向量的欧式距离,获得到消除阴影后的最终图像。实验表明,即使对于不同种类图像,该方法也可以获得很好的阴影增强效果。
This paper presents a new method for eliminating shadowed digital image combining high-dimension geometry calculation. In order to calculate a new image based on original one, it gets the new image through the way of convolution original image with a set of smoothing filter kernels, and then calculates the distance in Euclidean space and gets the last image point. Experimental results have shown there are better effects for shadowed image even for different category pictures.
出处
《计算机工程》
CAS
CSCD
北大核心
2007年第12期202-204,共3页
Computer Engineering
关键词
高维空间几何计算
图像消影
图像增强
High-dimension space geometry
Shadows elimination
Image enhancement
