面向服装信息识别的数据分类框架研究

Research on Data Classification Framework for Clothing Information Recognition

当前大部分消费者在衣着方面的消费都是基于线上进行,线上服装销售可以做到更全面、更多样化的推荐。针对虚拟试衣镜APP的服装信息分类需求,本文将面向服装数据的分类和规范化整理,为用户提供更加方便快捷的线上购物,提升用户的消费体验。通过线上的多种服装销售渠道进行服装信息的收集;根据服装信息中的标签信息对服装进行类别的确定;进一步对服装数据进行规范化,图片统一格式、标签信息按照规范化规则进行录入、尺码信息保留原始数据中的三个尺码并按照特定顺序将尺码中的数据信息录入,需要严格按照一定的规范化规则。最终经过规范化处理后的数据,可以直接提供给分类算法以便对服装信息的快速识别。

Nowadays,most of the consumers'clothing consumption is online.Online clothing sales can achieve m ore comprehensive and diversified recommendations.According to the needs of clothing information classificati on of virtual fitting mirror app,this paper will classify and standardize the clothing data to provide users with mo re convenient and fast online shopping and improve the user's consumption experience.Collect garment informat ion through various online garment sales channels;determine the category of garment according to the label infor mation in the garment information;further standardize the garment data,unify the format of pictures,input the la bel information according to the normalization rules,retain the three sizes in the original data and record the data information in the size in a specific order In the process of standardization,we need to strictly follow certain sta ndardization rules.Finally,the standardized data can be directly provided to the classification algorithm for the ra pid recognition of clothing information.