Shape recognition system based on neural networks and its optimization

It is not a new thing to use Hough transform in shape recognition or neural networks in classification. In this paper, combining the two methods together, we use the neural networks to realize the Hough transform and feature extraction. The feature extraction is an 1-D vector and is not affected by the sizes and positions of extracting objects. With the help of the shape recognition systems of constructed neural networks, a very good experiment result is obtained by using the BP algorithm to carry out studying, training and optimizing system parameters.