一种图像分割算法的选择方法

An Approach to the Selection of Image Segmentation Algorithms

提出了一种基于模糊联想分类神经网络图像分割算法的选择方法 .通过树型小波变换提取图像特征 ,再用模糊联想分类神经网络将待分割图像划分为灰度图像或纹理图像 。

An approach to image segmentation algorithm selection is presented. First, features of the image to be segmented are extracted by the tree structured wavelet transform. Then a fuzzy associative classification neural network is used to classify the image into gray area dominated image or texture area dominated one,which are used to choose an appropriate image segmentation algorithm. Experimental results show the effectiveness of the proposed approach.