Abstract:

In view of the problem of digital inheritance and innovation for blue calico, a method to extract image elements for blue calico based on contour fitting was presented. These image elements were the material foundation of blue calico to design and innovation. Firstly, the acquired blue calico digital images were preprocessed, including graying, denoising and binarization, to obtain a binary image reflecting the features of the image elements. Secondly, the initial contours of image elements were obtained by using contour tracking method, and the contours were fitted and optimized to get the ideal contours of image elements. Finally, the image elements of the blue calico were extracted and the sub images of image elements were generated, and they can be accessed independently. The experimental results show that using the evaluation indexes based on the area and contour pixels, the extracted image elements are accurate, and the contours are smooth and complete. More effective contours of image elements and execution efficiency are obtained than the other contour extraction methods.

Key words: blue calico, image element, contour extraction, evaluation index