Object Recognition with Lower Order Gegenbauer Moments
Simon Liao, Jing Chen · Lecture Notes on Software Engineering · 2013
Moment descriptors have been used as the image features in pattern recognition since the early development stage of moment method.The moment method is different from other object feature extraction methods as it captures global properties of an object rather than the local ones.In this research, an effective procedure to select moment features of an object is proposed.By using a set of 6,763 Chinese characters as the testing objects, we will show that the new Gegenbauer moment features perform very well in recognizing all Chinese characters including those that are very close in shapes.