A Theoretical Comparison of Texture Algorithms

Conners, Richard W.; Harlow, Charles A.

doi:10.1109/tpami.1980.4767008

Cited by 740 publications

(320 citation statements)

References 22 publications

Supporting

Mentioning

308

Contrasting

Unclassified

Order By: Relevance

“…In this method a segmentation algorithm is applied to a feature vector ®eld computed with the operator, and the segmentation performance and suitability of the used features are evaluated by using the number of misclassi®ed pixels (Weszka et al 1976;Conners and Harlow 1980;Pichler et al 1996). In this way, the practical value of the operator can be evaluated and compared with other texture operators.…”

Section: Classi®cation Results Comparison With Other Texture Operatorsmentioning

confidence: 99%

Computational model of dot-pattern selective cells

Kruizinga

Petkov

2000

Biological Cybernetics

View full text Add to dashboard Cite

Section: Classi®cation Results Comparison With Other Texture Operatorsmentioning

confidence: 99%

Computational model of dot-pattern selective cells

Kruizinga

Petkov

2000

Biological Cybernetics

View full text Add to dashboard Cite

“…8, minimum and total variations are not su cient to discriminate between all textures [42][43][44] (as they are very similar to the contrast [45] and the gray level di erence statistics [46]), but it is recognized that in general, di erent operators correspond to di erent images. In this work, speciÿc knowledge of the acquisition process is not applied and we try to demonstrate the validity of our neural implementation to a greater extent than the operator's one.…”

Section: Comparisons With Other Segmentation Methodsmentioning

confidence: 99%

A clustering fuzzy approach for image segmentation

Cinque

Foresti

Lombardi³

2004

Pattern Recognition

View full text Add to dashboard Cite

Segmentation is a fundamental step in image description or classiÿcation. In recent years, several computational models have been used to implement segmentation methods but without establishing a single analytic solution. However, the intrinsic properties of neural networks make them an interesting approach, despite some measure of ine ciency. This paper presents a clustering approach for image segmentation based on a modiÿed fuzzy approach for image segmentation (ART) model. The goal of the proposed approach is to ÿnd a simple model able to instance a prototype for each cluster avoiding complex post-processing phases. Results and comparisons with other similar models presented in the literature (like self-organizing maps and original fuzzy ART) are also discussed. Qualitative and quantitative evaluations conÿrm the validity of the approach proposed.

show abstract

“…In general, the PSM has been shown to be much less efficient than most other texture analysis methods [5,31]. Although Jernigan et al [14,17] propose to use entropy, peak, and shape measures to extract more texture features from the power spectrum, the performance improvement is limited, and the method is not widely accepted as an efficient texture analysis algorithm.…”

Section: Fourier Transform Featuresmentioning

confidence: 99%

“…Historically, the texture classification performance of the PSM has been ranked fairly low among most texture analysis techniques [5,31], resulting in limited application of this approach. Criticism of the PSM have been of the Fourier transform rather than of the way that texture features are computed from the power spectrum [31].…”

Section: Introductionmentioning

confidence: 99%

Optical and Sonar Image Classification: Wavelet Packet Transform vs Fourier Transform

Tang

Stewart

2000

Computer Vision and Image Understanding

View full text Add to dashboard Cite

To develop a noise-insensitive texture classification algorithm for both optical and underwater sidescan sonar images, we study the multichannel texture classification algorithm that uses the wavelet packet transform and Fourier transform. The approach uses a multilevel dominant eigenvector estimation algorithm and statistical distance measures to combine and select frequency channel features of greater discriminatory power. Consistently better performance of the higher level wavelet packet decompositions over those of lower levels suggests that the Fourier transform features, which may be considered as one of the highest possible levels of multichannel decomposition, may contain more texture information for classification than the wavelet transform features. Classification performance comparisons using a set of sixteen Vistex texture images with several level of white noise added and two sets of sidescan sonar images support this conclusion. The new dominant Fourier transform features are also shown to perform much better than the traditional power spectrum method.

show abstract

A Theoretical Comparison of Texture Algorithms

Cited by 740 publications

References 22 publications

Computational model of dot-pattern selective cells

Computational model of dot-pattern selective cells

A clustering fuzzy approach for image segmentation

Optical and Sonar Image Classification: Wavelet Packet Transform vs Fourier Transform

Contact Info

Product

Resources

About