作者: Fausto Espinal
DOI: 10.1117/1.601844
关键词: Pattern recognition 、 Contextual image classification 、 Automatic target recognition 、 Wavelet transform 、 Computer vision 、 Artificial intelligence 、 Wavelet 、 Fractal 、 Image segmentation 、 Feature extraction 、 Computer science 、 Fractal analysis
摘要: Texture measures offer a means of detecting targets in back- ground clutter that has similar spectral characteristics. Our previous studies demonstrated the ''fractal signature'' (a feature set based on fractal surface area function) is very accurate and robust for gray- scale texture classification. This paper introduces new multichannel model characterizes patterns as 2-D functions Besov space. The wavelet-based signature generates an n-dimensional surface, which used Results some experimental are presented demonstrating usefulness this mea- sure. © 1998 Society Photo-Optical Instrumentation Engineers. (S0091-3286(98)01001-0)