An Adaptive Texture Segmentation Method Based on Property of HVS
基于人类视觉特性的纹理分割方法

Texture segmentation involves subdividing an image into differently textured regions. Many texture-segmentation schemes are based on a filter-bank model, e.g. Gabor filter-bank. But these algorithms usually are limited because of the texture's variety and the parameters selection difficulty. These deficiencies lead to the inefficient extracted feature, the worse classification results and the limited application domain. In this paper, a texture segmentation algorithm based on the HVS mechanism was presented. It can give the number of texture type and approximate areas in the image mimicking the preattentive stage, and can also give the Gabor filters parameters mimicking the focused attention stage. This algorithm corresponds to the HVS property, and the approach is easy to implement and computationally efficient. The features are characteristic, and the classification results have higher accuracy.