作者: S. Levialdi
DOI: 10.1007/978-3-642-82017-5_1
关键词: Viewpoints 、 Theoretical computer science 、 Machine vision 、 Computation 、 ILLIAC 、 Von Neumann architecture 、 Computational problem 、 Digital image processing 、 Computer science 、 Image processing
摘要: The goal of a great part the research activity in digital image processing is twofold: 1) to design and efficiently implement algorithms for analysis description 2) improve understanding human visual perception system. An old discussion about usefulness these two aspects -and their complementarity -is still alive since some people, ways which Nature has established perform vision should have no bearing on an artificial system be designed. Conversely, group having opposite view claims that evolution significant process optimized solution computational problem such as one posed by (in 3-D, with colour and, instances, motion). As examples different thinking, first may represented work done M.L. Minsky, especially geometry(1) another example mathematical morphology (2) J. Serra , whilst second one, namely biologically based approach well (3) depicted B. McCormick starting early days ILLIAC computers up interesting proposal highly parallel structure real time processing(4). three cultural areas been mentioned reflect important viewpoints: geometry discrete needs special mathematics, set theory can used define picture algebra good knowledge physiology redesigning computer architectures will computation at levels are close ones natural systems. reason mentioning approaches they all share common point: traditional Von Neumann architecture not adequate operations neither reasonable efficiency nor level flexibility similar