Principles of organization selection and preprocessing video information. Bioinspired approaches

Expansion application branches of computer vision systems and requirements to them causes the necessities of significant increase of their productivity, efficiency and level of intelligence. The essential limitation of traditional video systems is the semantic gap in how person and video camera perceive the image and the information-time gap between parallel light perception and subsequent pixel processing. The prototype for creating more perfect computer vision systems can be a human visual analyzer. The paper proposes a number of approaches and methods of selective selection and multi-layer processing of video information, which reduce or eliminate these limitations, inspired by the human visual analyzer. This will increase the productivity, efficiency and intellectual level of computer vision systems.