An adaptive approach to unsupervised texture segmentation using M-Band wavelet transform

• Published in: Signal processing Vol. 81; no. 7; pp. 1337 - 1356
• Main Authors: Acharyya, Mausumi, Kundu, Malay K.
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 01-07-2001; Elsevier Science
• Subjects: Applied sciences; Exact sciences and technology; Feature extraction; Image processing; Information, signal and communications theory; M-band wavelets; Multiscale representation; Signal processing; Telecommunications and information theory; Texture segmentation; M-band wavelets; Feature extraction; Multiscale representation; Texture segmentation; Performance evaluation; Image texture; Texture analysis; Automatic classification; Subband decomposition; Image processing; Energy measurement; Linear phase; Adaptive method; Non linear transformation; Image segmentation; Statistical test; Wavelet transformation; Multiscale method
• ISSN: 0165-1684; 1872-7557
• DOI: 10.1016/S0165-1684(00)00278-4

The
M-band wavelet decomposition, which is a direct generalization of the standard 2-band wavelet decomposition is applied to the problem of an unsupervised segmentation of two texture images. Orthogonal and linear phase
M-band wavelet transform is used to decompose the image into
M×
M channels. Various combinations of these bandpass sections are taken to obtain different scales and orientations in the frequency plane. Texture features are obtained by subjecting each bandpass section to a nonlinear transformation and computing the measure of energy in a window around each pixel of the filtered texture images. The window size in turn is adaptively selected depending on the frequency content of the images. Unsupervised texture segmentation is obtained by simple
K-means clustering. Statistical tests are used to evaluate the average performance of features extracted from the decomposed subbands.