Unsupervised texture segmentation via wavelet transform

Chun Shien Lu, Pau-Choo Chung, Chih Fan Chen

Research output: Contribution to journalArticle

139 Citations (Scopus)

Abstract

In this paper, a mechanism for unsupervised texture segmentation is presented. The approach is based on the multiscale representation of the discrete (dyadic) wavelet transform which can be implemented by a fast iterative algorithm. For unsupervised segmentation it is generally difficult to determine the number of classes to be identified. The proposed approach offers an approach to circumvent this problem. Our method utilizes a set of high-frequency channel energies to characterize texture features, followed by a multi-thresholding technique for coarse segmentation. The coarsely segmented results at the same scale are incorporated by an intra-scale fusion procedure. A fine segmentation technique is then used to reclassify the ambiguously labeled pixels generated from the intra-scale fusion step. Finally, the number of texture classes is determined by an inter-scale fusion in which the segmentation results at multiple scales are integrated. The performance of this method is demonstrated by several experiments on synthetic images, natural textures from Brodatz's album and real-world textured images. Since the choice of wavelets is very extensive and open, we further explore various types of wavelets for texture segmentation. The time cost of the proposed method is also measured.

Original languageEnglish
Pages (from-to)729-742
Number of pages14
JournalPattern Recognition
Volume30
Issue number5
DOIs
Publication statusPublished - 1997 Jan 1

Fingerprint

Wavelet transforms
Textures
Fusion reactions
Discrete wavelet transforms
Pixels
Costs
Experiments

All Science Journal Classification (ASJC) codes

  • Software
  • Signal Processing
  • Computer Vision and Pattern Recognition
  • Artificial Intelligence

Cite this

Lu, Chun Shien ; Chung, Pau-Choo ; Chen, Chih Fan. / Unsupervised texture segmentation via wavelet transform. In: Pattern Recognition. 1997 ; Vol. 30, No. 5. pp. 729-742.
@article{1448c46c13a14814b5d4d1f8e65862b6,
title = "Unsupervised texture segmentation via wavelet transform",
abstract = "In this paper, a mechanism for unsupervised texture segmentation is presented. The approach is based on the multiscale representation of the discrete (dyadic) wavelet transform which can be implemented by a fast iterative algorithm. For unsupervised segmentation it is generally difficult to determine the number of classes to be identified. The proposed approach offers an approach to circumvent this problem. Our method utilizes a set of high-frequency channel energies to characterize texture features, followed by a multi-thresholding technique for coarse segmentation. The coarsely segmented results at the same scale are incorporated by an intra-scale fusion procedure. A fine segmentation technique is then used to reclassify the ambiguously labeled pixels generated from the intra-scale fusion step. Finally, the number of texture classes is determined by an inter-scale fusion in which the segmentation results at multiple scales are integrated. The performance of this method is demonstrated by several experiments on synthetic images, natural textures from Brodatz's album and real-world textured images. Since the choice of wavelets is very extensive and open, we further explore various types of wavelets for texture segmentation. The time cost of the proposed method is also measured.",
author = "Lu, {Chun Shien} and Pau-Choo Chung and Chen, {Chih Fan}",
year = "1997",
month = "1",
day = "1",
doi = "10.1016/S0031-3203(96)00116-1",
language = "English",
volume = "30",
pages = "729--742",
journal = "Pattern Recognition",
issn = "0031-3203",
publisher = "Elsevier Limited",
number = "5",

}

Unsupervised texture segmentation via wavelet transform. / Lu, Chun Shien; Chung, Pau-Choo; Chen, Chih Fan.

In: Pattern Recognition, Vol. 30, No. 5, 01.01.1997, p. 729-742.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Unsupervised texture segmentation via wavelet transform

AU - Lu, Chun Shien

AU - Chung, Pau-Choo

AU - Chen, Chih Fan

PY - 1997/1/1

Y1 - 1997/1/1

N2 - In this paper, a mechanism for unsupervised texture segmentation is presented. The approach is based on the multiscale representation of the discrete (dyadic) wavelet transform which can be implemented by a fast iterative algorithm. For unsupervised segmentation it is generally difficult to determine the number of classes to be identified. The proposed approach offers an approach to circumvent this problem. Our method utilizes a set of high-frequency channel energies to characterize texture features, followed by a multi-thresholding technique for coarse segmentation. The coarsely segmented results at the same scale are incorporated by an intra-scale fusion procedure. A fine segmentation technique is then used to reclassify the ambiguously labeled pixels generated from the intra-scale fusion step. Finally, the number of texture classes is determined by an inter-scale fusion in which the segmentation results at multiple scales are integrated. The performance of this method is demonstrated by several experiments on synthetic images, natural textures from Brodatz's album and real-world textured images. Since the choice of wavelets is very extensive and open, we further explore various types of wavelets for texture segmentation. The time cost of the proposed method is also measured.

AB - In this paper, a mechanism for unsupervised texture segmentation is presented. The approach is based on the multiscale representation of the discrete (dyadic) wavelet transform which can be implemented by a fast iterative algorithm. For unsupervised segmentation it is generally difficult to determine the number of classes to be identified. The proposed approach offers an approach to circumvent this problem. Our method utilizes a set of high-frequency channel energies to characterize texture features, followed by a multi-thresholding technique for coarse segmentation. The coarsely segmented results at the same scale are incorporated by an intra-scale fusion procedure. A fine segmentation technique is then used to reclassify the ambiguously labeled pixels generated from the intra-scale fusion step. Finally, the number of texture classes is determined by an inter-scale fusion in which the segmentation results at multiple scales are integrated. The performance of this method is demonstrated by several experiments on synthetic images, natural textures from Brodatz's album and real-world textured images. Since the choice of wavelets is very extensive and open, we further explore various types of wavelets for texture segmentation. The time cost of the proposed method is also measured.

UR - http://www.scopus.com/inward/record.url?scp=0031143685&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0031143685&partnerID=8YFLogxK

U2 - 10.1016/S0031-3203(96)00116-1

DO - 10.1016/S0031-3203(96)00116-1

M3 - Article

AN - SCOPUS:0031143685

VL - 30

SP - 729

EP - 742

JO - Pattern Recognition

JF - Pattern Recognition

SN - 0031-3203

IS - 5

ER -