Neural network based image recognition system using geometrical moment

摘要

Geometrical moments (GM) have been used in the classification of four closed planar shapes (Gupta and Srinath, 1987). Also a neural network approach for the classification of four closed planar shapes has been used in Khotanzad and Lu (1990). In this paper, a backpropagation neural network is used in the recognition of six different kinds of hand tools using geometrical moments. Experimental results indicate that the neural network approach gives a better recognition accuracy when compared with the two conventional statistical classifiers-namely the single nearest neighbour and minimum-mean-distance. Recognition accuracy using a neural network is over 98%.