Modeling of scaled and rotated images recognition using different variants of invariant correlation filters

摘要

The presented paper deals to PC modeling of scaled and rotated images recognition based on different types of distortion invariant correlation filters. There was used a database of different images (both false and true class) that are under geometrical distortions and there were calculated correlation outputs for images recognition with the help of MACE, MINACE, GMACE, DCCF and polynomial filters and it's combinations for two-stage recognition. The results provide a possibility for successful usage of implemented algorithms.