This paper proposes a novel texture feature extraction technique for texture image retrieval. The method is robust to geometric distortions as well as noise effect. The geometric distortions include rotation, scaling and translation modifications of textures. In the feature extracting process, log-polar transformed autocorrelation images are introduced to eliminate the effects of the entire distortions. The influence of additive noise is reduced by modifying autocorrelation images. In the retrieval process, valuable wavelet packet statistics is used to measure similarity between individual images. The effectiveness of our method is demonstrated using noisy distorted texture image database in the experimental simulations.
Robust feature extraction technique for texture image retrieval
2005-01-01
230034 byte
Conference paper
Electronic Resource
English
Robust Feature Extraction Technique for Texture Image Retrieval
British Library Conference Proceedings | 2005
|Rotation-invariant texture feature for image retrieval
British Library Online Contents | 2003
|Regular-texture image retrieval based on texture-primitive extraction
British Library Online Contents | 1999
|Robust-to-rotation texture descriptor for image retrieval in wavelets domain
British Library Online Contents | 2006
|British Library Online Contents | 2012
|