Digital Rebirth of Dongba Pattern: An Improved Active Contour Model for Pattern Contour Extraction

Abstract

In response to the time-consuming and laborious problem of manually drawing elements in existing poster designs, this paper proposes a method for extracting the contours of Dongba patterns based on improved active contour model. The improved active contour method adds discrete wavelet transform for energy minimization. Moreover, the minimization formula assigns greater weight to foreground pixels. In addition, to prevent evolutionary instability during the iteration process, we employed an optimization process that is specifically designed to maintain clear contours along the edges of the pattern. Based on the established Dongba pattern dataset, the qualitative and quantitative analyses are conducted on the proposed method. The results indicate that the proposed improved active contour model has advantages in contour extraction visual quality and evaluation indicators compared to existing methods, demonstrating the effectiveness of the proposed method. Finally, based on the proposed pattern contour extraction method, the design innovation practice of Dongba cultural posters is completed from two aspects: pattern selection and transformation, artistic processing and poster design. The code is available at: https://github.com/jsluen/IACM.