Pahomov A.N.

The problem - orientational yild of information jointly with the higlilevel control of eye movements are allowed intellect all numerous of environment world. Due to visual system of artificial intellect this signify that it must satisfy to olifferent kind of activity such as pursuit, watch, navigation and etc (etcetera) . According o this the active videosensor must to realize such transformation of input signal image that provide the objectivity and constantivity of reseption. Many world were devoted to the active vision system and, particulary, to active video sensors. Analisis this works is shown that realization of problem-orientational yild of information often carry out by means of input image processing in state-space region. More over sufficiently often the processing is ensured with use the directive image shifts concerning light sensitive surface of sensor that like non-controlled movements of human eye. At the same time is observerd the understatement of the image processing in freguency region in the vision analizator. Confirmation of wide possibilities of image processing in frequency region are the result of long standing research of discret-analog formcalculational mediua. One from record conclusions consist of dimension spatio-frequency filters significantly widen in case of image shift concerning light-sensitive surface of sensor. Furthermore the kind of route transducer movement is defined the view of filter amplitude-frequency performance. The results fresearchs are allowed to draw the conclusions: the attempts to solve all complex tasks on realization of active perception mechanism based on algorithm of processing of spatial region, probably in the near future will not carried success; non-controlled eye movements are provided the realization problem-orientate perception by optic flow processing as in spatial region as in frequency region; hence this way must be definitional at active video-sensors implementation. Slow eye drift may play one from the important roles at the frequency processing and defined the track pass and accordingly to turning of the spatial-frequency filters of the visual analyzator.