Connecting and fitting of curve segments based on the genetic algorithm
Hailin Li · Jounal of Xidian University · 2000
A feature filter is presented to extract linear points and their relations from input samples. Then, curve segments are obtained by classifying the samples on the basis of their relations. Finally, the detection and recognition of curves are completed by the combination of the Genetic Algorithm and fitting procedures. Experimental results show that the presented method not only succeeds in the recognition task but also is robust to noise.