A new approach for rotation-invariant and noise-resistant texture analysis and classification |
| |
Authors: | " target="_blank">Mohammad Mahdi Feraidooni Davood Gharavian |
| |
Affiliation: | 1.Department of Telecommunications, Faculty of Electrical Engineering,Shahid Beheshti University, G. C.,Tehran,Iran |
| |
Abstract: | The analysis and classification of images, such as texture images, is one of the substantial and important fields in image processing. Due to destructive effects of image rotation and noise, the stability and efficiency of texture analysis and classification methods are an important research area. In this paper, a new method for texture analysis and classification has been proposed which is based on a particular combination of wavelet, ridgelet and Fourier transforms as well as support vector machine. The proposed method has been evaluated for 13 texture datasets produced by three original datasets containing 25 and 111 original textures from Brodatz database and 24 original textures from OUTEX database. These datasets comprise 415584 and 93600 rotated noise-free and noisy texture images for Brodatz database and also 49920 noisy and 4320 noise-free texture images for OUTEX database, respectively. Simulation results demonstrate the capability, efficiency and also stability of the proposed method especially for real-time rotation-invariant and noise-resistant texture analysis and classification. |
| |
Keywords: | |
本文献已被 SpringerLink 等数据库收录! |
|