基于商空间的纹理图像分割

Texture Image Segmentation Based on Quotient Space

商空间理论是新近兴起的基于人工智能研究领域的一个很有潜力的方向。文中将商空间理论应用于纹理图像的分析 ,通过对纹理的区域结构特征分析研究 ,提出基于象素 8 邻域周期排列的纹理区域特征 ,并对结构性多纹理图像进行了分割实验 ,效果令人满意。文章总结了运用商空间理论求解复杂问题一般过程 。

Theory of quotient space is a potential direction about research field based on artificial intelligence in recent years. The theory of quotient space is applied to the analysis of texture image, and the texture feature is extracted from 8 neighborhood pixel cycle form, which would segment the texture image by analyzing the field features of texture image, and the test result of structural multi texture image is satisfactory. In the end, the paper concludes the process, during which the quotient space theory is applied to work out the complicated questions, and this paper also demonstrates the practical value of the theory.