ABSTRACT
Creating effective data analysis with efficient representation of complex hierarchical architectures is one of the key issues when developing novel computer systems. This chapter presents the multistage methodology applied to image processing. The concept of parallel multistage image processing can be considered as a consecutive transformation of image components, which consist of elements grouped under a certain criterion. It offers a new approach to the creation of computing medium, parallel-hierarchical (PH) networks, investigated in the form of a model of a neuro-like scheme of data processing. The main advantage of this approach is using the dynamics of multilevel parallel interaction of information signals at different hierarchy levels of computer networks that allows to use such known natural features of computation organisation in the cortex as: topographic nature of mapping, simultaneity (parallelism) of signal operation, inlaid cortex, structure, rough hierarchy of the cortex, or the mechanism of perception and training spatially correlated in time.
