从服装图片中识别面料方法的研究

提出了一种服装面料自动识别方法:根据需要对服装图片进行预处理,然后从预处理后的图像中提取服装面料的特征信息,并对这些信息进行仿真,观察模拟效果,最后对从图像中识别出的面料信息进行分析,总结出几种服装面料图像的参数分布。对服装面料自动识别技术具有一定的推动作用,对虚拟现实中具有随机几何模型的物体特征识别也有一定的推动作用。

Research on Clothes Fabric Specification from Fashion Images

A clothes fabric recognition method is presented in this study to automatically extract and specify its fabric type features. Clothes image-preprocess is firstly employed according to the different cases. Based upon the results of such image-processes, the clothes fabric feature information can be efficiently extracted. Secondly, a series of measures including statisticing, arranging and anglicizing of such information is correspondingly taken. Through simulating and redisplaying these feature information, the features of parameter distribution may be finally achieved, based on which the clothes fabric type is specified exactly. With the fast development of fashion manufacturing technology, such method, presented by this study, will play an active promoting role not only in clothes fabric recognition and specification, but also in feature recognition of the objects with stochastic geometric model in virtual reality field.