Shapes matching based on chain code and dynamic time warping

In this research, we indicate the use of Freeman Chain Code (FCC) method in image representation and focus on the shapes chain code as the feature extraction for shape matching. Each shape has a different vector in length of chain code hence it needs to match with the template image of the chain code. By using Dynamic Time Warping (DTW) for classification; we manage to indicate the distance between the input images and template image. The lowest distance shows that the image is the closest match. We tested on various regular and irregular shapes, and the result of the shape matching on regular shapes is precise and acceptable but for irregular shapes, some of the dissimilar shapes have almost the same distance and it's invalid.