RECOGNITION OF LINE-SHAPES USING NEURAL NETWORKS

We apply neural networks to implement a line shape recognition/classif ication system. The purpose of employing neural networks is to elimina te target-specific algorithms from the system and to simplify the syst em. The system needs only to be trained by samples. The shapes are cap tured by the following operations. Lines to be processed are segmented at inflection points. Each segment is extended from both ends of it i n a certain percentage. The shape of each extended segment is captured as an approximate curvature. Curvature sequence is normalized by size in order to get a scale-invariant measure. Feeding this normalized cu rvature data to a neural network leads to position-, rotation-, and sc ale-invariant line shape recognition. According to our experiments, al most 100% recognition rates are achieved against 5% random modificatio n and 50% - 200% scaling. The experimental results show that our metho d is effective. In addition, since this method captures shape locally, partial lines (caused by overlapping etc.) can also be recognized.