摘要
针对目前服装定制系统未能依据用户个人体貌特征给出服装推荐建议的现状,提出了一种基于三维人脸特征的服装个性化推荐及虚拟展示方法.采用主动外观模型提取正面人脸图像的关键特征点,并基于三维变形模型生成客户的个性化三维人脸.定义了客户脸型和肤色划分及计算方法,依据客户体型、脸型和肤色等体貌特征,结合文中给出的服装推荐规则,实现了服装的个性化推荐.基于Unity3D平台,实现了基于三维人脸特征的服装个性化推荐及虚拟展示原型系统,客户还可以在线与系统交互,自主选择服装款式及面料,系统实时展示服装着装效果,为服装个性化定制提供了理论和技术支持.
At present, clothing customization system has almost no recommendation of clothing according to the characteristics of users′ physical appearance. A clothing recommendation and virtual display approach based on 3D face modeling features is presented in this paper. The Active Appearance Models is used to extract the facial landmark points from customers′ frontal facial images. Then the individual 3D face model of the customer is reconstructed based on these key points and 3D morphable models. The method of calculating the customers′ face shape and skin color is defined in the article. According to the physical features of customers′ body shape, face shape and skin color, the matching clothing recommendation rules are established. In addition, the recommendation rules are given. The prototype system of personalized recommendation and virtual display about 3D face modeling features are developed based on Unity 3D platform. Customers can choose the clothing style and fabric through the system interaction, and the system could show the clothing effect in real time. It would provide theoretical and technical support for personalized recommendation, customization and virtual display of clothing.
作者
吕海清
朱欣娟
赵雪青
石美红
LYU Haiqing;ZHU Xinjuan;ZHAO Xueqing;SHI Meihong(School of Computer Science/Shaanxi Key Laboratory of Clothing Intelligence,Xi′an Polytechnic University,Xi′an 710048,China)
出处
《纺织高校基础科学学报》
CAS
2018年第3期317-325,共9页
Basic Sciences Journal of Textile Universities
基金
国家自然科学基金(61806160)
关键词
个性化服装定制
个性化推荐
交互式服装设计
服装虚拟展示
personalized clothing customization
personalized recommendation
personalized 3D face
clothing virtual display
