Digital simulation of matched pattern detection by means of computer-generated spatial filter

The matched pattern recognition for the two-dimensional image has been performed by the optical and/or the digital techniques. The digital simulation techniques for the matched filtering by means of a computer-generated spatial filter have more or less problems in point of detectable accuracy of the pattern in comparison with the results of the coherent optical setups. In this paper, the author demonstrates ths use of modified spatial filter such as a high pass filter, a differential filter, a Laplacian filter and a S/N matched filter instead of a general spatial filter in order to improve detectable accuracy of the pattern. It is possible to accurately carry out the intuitive and quantitative comparison of results of the matched pattern by the digital correlation techniques.