Feature Extraction for Texture classification – An Approach with Discrete Wavelet Transform
- A M. Kothari, Atmiya inst of science and technology, Rajkot, Gujarat, INDIA
- D G. Kamdar, VVP Engineering College, Rajkot, Gujarat, INDIA, .
- D. A. Doshi, Atmiya Institute of Tech. and Science, Rajkot, Gujarat, INDIA, firstname.lastname@example.org
Texture is used in many areas such as remote sensing, surface detection, biomedical image processing etc. Surface metrology with image processing is a challenging task having wide applications in industry. Surface roughness can be evaluated using texture classification approach. In this paper concentrated on Multi-scale multi dimensional technique for classification. Features which are included for classification of surface are Wavelet Statistical Features (WSF) and Wavelet co-occurrence Features (WCF). Important aspect here is appropriate selection of features that characterize the surface. Multi-directional transform for fast and robust feature extraction, where scale and angular decomposition properties are integrated to increase its texture classification performance. Here DWT is used for detecting surface feature. By selecting appropriate features, classification rate can be enhanced.