Determination of the number of texture segments using wavelets

Abstract

This paper presents a robust method of determining the number of texture segments in an image. We take an $N imes N$ image and decompose it into $n imes n$ blocks. A three-scale two-dimensional discrete wavelet transform is performed on each $n imes n$ block. For each $n imes n$ 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.

Keywords

• Differential geometry
• algebraic geometry