Research On Intelligent Measurement And Recommendation Technology For Suit Customization

With the rapid development of the clothing industry in e-commerce,buying clothing online has become a popular choice.However,the lack of effective online measurement methods for suit customization services limits its development in e-commerce.Therefore,a low-cost and convenient measurement technology method is urgently needed to realize online body measurement.The research in this thesis proposes a non-contact body measurement method based on BodyPix,which provides a new solution for online body size measurement.At the same time,considering the requirements of suits for fit,a suit recommendation system based on human body characteristics combined with an expert system is proposed.Through the deep image processing of the human body,the customer’s body shape data is obtained,and the expert system is used to implement intelligent suit recommendation for users,allowing users Have a better experience when shopping.This article mainly completes the following three aspects of work:1.Apply BodyPix to the field of human body measurement,based on the human body segmentation image obtained by BodyPix,and realize the non-contact body measurement technology based on BodyPix through image processing technology and body size fitting method.2.In order to realize the intelligent recommendation of suits based on the contour features of the human body combined with the expert system,the contour image of the human body is first extracted.With the popularization of the application of depth lens on mobile phones in the future and the needs of on-site measurement in stores and group outfits,this thesis proposes a depth camera based RealSense D435 obtains the depth image of the human body,combined with the improved Canny edge detection algorithm,the method of realized according to the human body contour to quickly determine the purpose of the human body,and provides a data basis for the suit intelligent recommendation system.3.The suit recommendation system is further realized on the basis of quickly judging the figure of the human body by acquiring the contour of the human body based on the depth image.In the early stage of the system,the body shape data and the data in the rule library in the expert system are used to generate recommendation rules to recommend suits matching the body shape for the current user.When the system has a large number of users,the user similarity is calculated according to the body shape(body shape,shoulder shape,face shape,height),and a new suit recommendation list is formed for the user by referring to the purchase records of similar users and the knowledge of the expert rule database.Users can also recommend suitable clothing fabrics for them by selecting the season,so that customers can experience a more scientific and intelligent recommendation service.