基于Unit-Linking PCNN和HSI空间的彩色图像分割方法

提出了一种基于Unit-Linking PCNN和HSI空间的彩色图像分割新方法.在RGB空间中,由于R,G,B分量之间有很高的相关性,直接利用这些分量不能得到所需的效果.在提出的算法中,彩色图像首先被转化到HSI颜色空间,图像分解成彩色通道(H和S分量)和亮度通道(I分量).对彩色通道采用基于Unit-Link-ing PCNN的分割算法进行分割,对亮度通道采用最大类间方差阈值算法进行分割,然后将分割后的结果用大概率合并的方法进行组合,得到最终分割结果.试验结果表明,这种方法具有很好的彩色图像分割效果.

Color image segmentation based on Unit-Linking PCNN and HSI space

This paper suggests a new method of color image segmentation based on Unit-Linking PCNN and HSIcolor space.In RGBspace,as a result of R,G,Bcomponents have a high relevance,the direct use of these components can not bethe desired result.In the algorithm of this paper,the color image is first transformed into HSIcolor space,image broken down into color channel(Hand Scomponent)and the brightness channel(Icomponent).We use Unit-Linking PCNN segmentation's algorithm for color channel and Otsu threshold segmentation algorithm on the luminance channel,and then after combination the segmentation's results using large probability,the final segmentation results were received.Experimental results show that such a method has preferable results of color image segmentation.