Abstract
Moment invariants based on image boundaries and silhouettes of objects are computed to form 14-dimensional feature vectors by using special functions provided by a new vision system. A quite general prototype training software package was developed on this system. Two different pattern classification approaches. Bayes decision approach and the distance-weighted- K-Nearest-Neighbor rule are implemented and compared. Experimental results and analysis of these results are given.
Full Text
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call