作者: Narendra Ahuja , Azriel Rosenfeld
DOI: 10.1016/S0169-7161(82)02020-3
关键词: Autocorrelation 、 Resolution (logic) 、 Pattern recognition 、 Computer vision 、 Computer science 、 Statistical model 、 Image texture 、 Variation (game tree) 、 Plane (geometry) 、 Structure (mathematical logic) 、 Artificial intelligence 、 Image (mathematics)
摘要: Publisher Summary This chapter discusses the models of spatial intensity variation in homogeneous images — that is, do not exhibit any macrostructure. The lack macrostructure results when a uniformly structured scene is photographed at coarse resolution, consequent dense packing structure giving rise to image texture. Traditionally, have been classified as statistical or structural. involve description statistics such autocorrelation, while structural approach specifies primitives and placement rules for laying these out plane. It should be noted if were statistical, resulting would too regular interest. If model cannot reveal basic an image, it powerful enough much use. pixel-based region-based are discussed reviews many studies modeling conducted through 1978.