Fundamental properties of the selected joint transform correlators for practical application in automatic images recognition

The paper presents review of the selected joined transform correlators (JTC) used for automatic images recognition. Because they are used in non-contact optical measurement systems thus special numerical processing should be used for analysis of data collected by them. The neutral network approach, wide development in last decade for such system, seems to be small useful, mainly because it creates computing model. The image is parallel information and such method as the Fourier transform are more useful for its processing. From this reason the wavelet transform as generalization of the Fourier method has been chosen as optimal. On this base the JTC uses wavelet transforms (WT) for analysis because it is a powerful tool for wavelet filtration, image feature extraction and recognition of rationally distorted images. The laboratory system uses the joint wavelet transform correlator (JWTC) and its possibilities were presented on this paper as final result of review.