作者: André R. Backes , Odemir M. Bruno
DOI: 10.1007/978-3-540-69905-7_16
关键词: Projective texture mapping 、 Fractal analysis 、 Texture filtering 、 Image texture 、 Texture (geology) 、 Artificial intelligence 、 Pattern recognition 、 Mathematics 、 Computer vision 、 Fractal dimension 、 Texture compression 、 Binary image
摘要: One of the most important visual attributes for image analysis and pattern recognition is texture. Its allows to describe identify different regions in through pixel organization, performing a better description classification. This paper presents novel approach texture analysis, based on calculation fractal dimension binary images generated from texture, using threshold values. The proposed performs complexity as values changes, producing signature which able characterize efficiently classes. illustrates method performance an experiment Brodatz images.