摘要
随着电商交易平台的飞速发展,服装行业的网上交易开始变得普及,但是网络服装行业的发展难以突破在销量上超过实体店的瓶颈,最主要的限制是网购的未知性和用户体验;没有用户的虚拟试穿体验,用户甚至不知道衣服的合身尺寸和穿上衣服的实际效果,而云端虚拟更衣室就是为了解决这一实际问题。该系统分为三个部分:C/S虚拟更衣室客户端、基于协同过滤的推荐算法后台框架、IIS服务器。该系统采用的辅助硬件为Kinect,图像处理算法采用Emgu来优化,前台客户端推荐列表采用基于协同过滤和基于内容过滤的联合推荐算法,匹配用户身高的衣服尺寸则采用Kinect SDK自带的骨骼追踪算法实现。项目设计的云平台虚拟更衣室是基于多算法融合的网络结构的框架,通过大量测试和实验表明,该项目有效地解决了体感应用与云平台结合的问题,并且经过完善后,用户体验更加友好。
With the rapid development of e-commerce trading platform,the apparel industry＇s online trading has become popular,but the development of online apparel industry is difficult to break through the bottle neck of the store,the most important is the unknown and user experience online shopping experience,users do not even know the actual effect of clothes and cloud virtual dressing room is to solve this problem. The system is divided into three parts： C/S virtual locker room client,based on collaborative filtering recommendation algorithm background framework,IIS server. The system uses the auxiliary hardware,which is Kinect,the image processing algorithm uses Emgu to optimize,the foreground client recommended list is based on collaborative filtering and content-based filtering recommendation algorithm,matching the user＇s height of the clothes size is the use of kinect SDK＇s own skeleton tracking algorithm. The project design of the cloud platform virtual dressing room is based on the framework of multi algorithm fusion of network structure,through a large number of test and experimental surface,the project effectively solves the problem of combining the application of the body and the cloud platform,and after improving,the user experience is more friendly.
出处
《信息技术》
2017年第1期39-43,共5页
Information Technology
基金
广东省大学生科技创新培育专项资金(攀登计划专项资金)(pdjh2015b0925)
北京理工大学珠海学院学生科研发展基金项目(2013XS10)
关键词
KINECT
云平台
图像处理算法
推荐算法
虚拟更衣室
Kinect
cloud platform
image processing algorithm
recommendation algorithm
virtual dressing room
