An in Depth Comparison of Four Texture Segmentation Methods
- 1 December 2007
- conference paper
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE) in 9th Biennial Conference of the Australian Pattern Recognition Society on Digital Image Computing Techniques and Applications (DICTA 2007)
Abstract
Texture segmentation is the process of partitioning an image into regions with different textures containing similar group of pixels. This paper presents a comparative study of four texture segmentation methods based on the following features: descriptors, heuristic function, fuzzy logic and Mask based features. Many types of textures are considered for analysis. The comparative results show that descriptor based approach is the most suitable for segmenting both natural and mosaic textures whereas heuristic function based approach is most suitable for random textures. Fuzzy features based approach is found to yield better segments for regular patterns while Mask feature based approach is the best for segmenting Natural images, but fails miserably on Mosaic textures. Fuzzy C-means classification is used for achieving texture segmentation.Keywords
This publication has 5 references indexed in Scilit:
- An Optimal Fuzzy System for Color Image EnhancementIEEE Transactions on Image Processing, 2006
- A fuzzy approach to texture segmentationPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2004
- An efficient approach to compute fractal dimension in texture imagePublished by Institute of Electrical and Electronics Engineers (IEEE) ,2003
- Texture segmentation using fractal dimensionIEEE Transactions on Pattern Analysis and Machine Intelligence, 1995
- Textural Features for Image ClassificationIEEE Transactions on Systems, Man, and Cybernetics, 1973