ABSTRACT

An assessment of productivity of the image processing system based on analysis of the function of the area of intersection (FAI) is carried out by the criterion of the quality of recognition. An important factor for each method of data processing is the selection of the removal of noise algorithm. The areas of the figures are analysed by methods of preprocessing and of image recognition on the basis of parallel shift technology. The image recognition system, based on the analysis of FAI, uses three types of permissible errors. The first kind of error is used when comparing the integral ratios of the FAI in the fast stage of the recognition process. The second kind of error is used for comparison of the FAI at a detailed stage of the recognition process. The third kind of error is used in the formula of value determination of maximum shifts.