作者: Alvaro G. Zuñiga , Joao B. Florindo , Odemir M. Bruno
DOI: 10.1016/J.PATREC.2013.09.023
关键词:
摘要: Texture analysis and classification remain as one of the biggest challenges for field computer vision pattern recognition. On this matter, Gabor wavelets has proven to be a useful technique characterize distinctive texture patterns. However, most approaches used extract descriptors magnitude space usually fail in representing adequately richness detail present into unique feature vector. In paper, we propose new method enhance process extracting fractal signature spaces. Each is reduced using canonical function concatenated form final Experiments were conducted on several image databases prove power effectiveness proposed method. Results obtained shown that outperforms other early method, creating more reliable extraction.