基于模糊理论的毛针织服装款式识别

Pattern recognition based on fuzzy theory for wool knitted clothing

提出了一种新的基于模糊理论的毛针织服装款式识别算法,使毛针织服装CAD更具智能化,可提高后续衣片拆分和工艺单计算的效率。该方法首先分析毛针织服装款式轮廓点的特点,建立了由7个特征组成的款式轮廓点分类依据;根据模糊理论建立了模糊聚类矩阵,通过计算机将款式轮廓点自动归类;根据毛针织服装款式的特点,给出了款式对比前的缩放公式,并提出了用于款式模糊识别的贴近度公式。得出了该套算法可快速准确将毛针织服装款式轮廓点归类并可准确识别款式类型的结论。

In this paper, a new pattern recognition algorithm for knitted clothing based on fuzzy theory was proposed. This algorithm made the knitted clothing CAD more intelligent and improved the efficiency of technological sheet calculation and clothing pieces disconnection. Firstly, the characteristic of clothing pattern silhouette point was analyzed, and the classifying criterion of pattern silhouette point made by seven characteristic points was constructed. Then the fuzzy clustering matrix based on fuzzy theory was established, and the pattern silhouette point was classified automatically by computer. At last, the shortening and enlarging formula based on knitted clothing pattern was presented, and the closeness function for pattern fuzzy recognition was put forward. In this paper, a conclusion was given that this algorithm classified the pattern silhouette point more quickly and exactly and the pattern type was recognized more correctly.