Determination of the Number of Texture Segments Using Wavelets

This paper presents a robust method of determining the number of texture segments in an image. We take an N ×N image and decompose it into n × n blocks. A three-scale two-dimensional discrete wavelet transform is performed on each n × n block. For each n × n block, this transformation produces coefficients for 25 wavelet channels. The energy of each channel is used as a tuple for a vector in the feature space. Nearest neighbor clustering is used to segment the feature space. A measure is defined to determine the “goodness” of the clustering. The optimal number of segments is taken to be the clustering which maximizes our measure.