Projection-invariant pattern recognition with logarithmic harmonic function and wavelet transform.
A logarithmic harmonic filter can detect objects at different projection angles. The Mexican-hat wavelet function can extract edges of equal width for objects, regardless of their sizes. Hence incorporating wavelet filtering in the logarithmic harmonic filter can improve its performance. The theory is presented together with computer simulation. Finally, an experiment using a joint transform correlator is presented to verify the capability of the proposed filter.
Cheng YS, Chen HC.
No user commented in " Projection-invariant pattern recognition with logarithmic harmonic function and wavelet transform. "
Follow-up comment rss or Leave a TrackbackLeave A Reply