| 摘要 |
automatics, computing technique, possibly technical vision systems for identifying images in robots of modern systems for finding, tracking, guiding, biomedical diagnostics and production process monitoring. SUBSTANCE: method comprises steps of converting parameters of images to logic-time function subjected to processing in quantity and quality channels; on base of such processing performing analysis and comparison of received date with reference sample data preliminarily stored in knowledge base memory. Apparatus for identifying images at eye- processor selecting of features includes unit for projecting images, optical lock, photoelectric converter in the form of matrix of MxM of photo- detecting cells, unit for converting image parameters to time interval duration values, unit for analysis of received data, unit for synthesizing- generating features, unit for recommutation, unit for generating switching logic-time function, dynamic memory unit for storing reference logic- time functions having system for organizing base of reference knowledge and selecting samples of reference data, comparator, unit for analysis of signal anticoincidence, image type shaper and driving pulse shaper. EFFECT: simplified process for image identifying, possibility for organizing evolution type knowledge base, enhanced responsibility and authenticity of identified images. 2 cl, 3 dwg _
|
