A transformation with invariance under cyclic permutation for applications in pattern recognition

摘要: The paper describes a transformation that can be used to characterize patterns independent of their position. Examples the application transform for machine recognition letters are discussed. program succeeded in rate 80–100% having different position, distortions, inclination, rotation up 15° and size variation 1:3 relative reference set 10 letters. Results with autonomous learning new varieties pattern (using matrix as an adaptive classifier) given. When executed on digital computer, this is 10–100 times faster than fast Fourier (depending number sampling points).