女装廓形精细化分类与智能识别

Refined Classification and Intelligent Identification of Women's Clothing Silhouette

对六大女装基础廓形(A形、H形、T形、O形、X形和S形)进行精细化分类和定义,并根据该分类对女装廓形进行智能识别。智能女装廓形识别的实现需从女装廓形基础数据库中调用基础尺寸数据,根据数据的不同组合将常见女装分成6种着装情况,通过计算服装各个主要部位之间的角度来反映服装的廓形曲线,用不同的角度区间来匹配对应不同的廓形分类,最终对女装廓形进行识别。将该方法应用于女装秀场图片中的廓形分析后,发现其具有较高的准确性,能为女装廓形趋势走向的智能识别与分析提供相关参考依据。

The refined classification and defining of six basic women＇s clothing silhouettes （A-line, H-line, T-line, O-line, X-line, and S-line） have been performed. And intelligent identification of the silhouette has been done based on the classification. The silhouette identification further classified the common women＇s clothing into six dressing situations following different data combinations. Then the clothing outline curves were captured by the angles among the characterizing body points. The six basic silhouette types were defined for all the six dressing situations considering different ranges of the calculated angles. Finally, the proposed methodology was implemented to identify the women＇s clothing silhouettes from the shows of various international brands. The result proves the accuracy of the proposed classification and identification. In addition, the proposed methodology provides some convincing references for the trend prediction of women＇s clothing silhouettes.