Jacquard Image Segmentation Method Based on Fuzzy Optimization Technology

Automatic pattern segmentation of jacquard images is a challenging task for jacquard pattern analysis. In this paper, the phase field model was introduced to extract specific pattern structures within jacquard images. A novel fuzzy optimization method, namely, Multi-start Fuzzy Optimization Method (MSFOM) was proposed for numerical solving of the phase field model. The proposed method was a hybrid algorithm combining fuzzy logic and genetic algorithms, which was able to find global minimum of the phase field model with low computational cost. Experimental results on synthetic and jacquard images demonstrate the efficacy of the proposed method.