The Design And Implementation Of Clothing Retrieval System Based On Mobile Phone

With the increasing maturity of the mobile internet, and the growing popularity of the mobile smart devices, people’s living standard is increasing, at the same time their economic activities are more and more diversified, as a result, e-commerce based on mobile platform has gained its development. Nowadays, personal mobile smart devices have varied e-commerce applications. People strive for the more convenient, more effective way of shopping on the mobile devices.This paper based on Android system, design and implement the content-based image retrieval system which aimed at the field of clothing, combines the image retrieval system with mobile smart phone. The main contents are as follows;1. Use the crawler system tools to collect domestic well-known e-commerce sites’ clothing images and related information. Research and solve the problem of URL duplication, extracting web information to set up a structured storage index database model.2. Based on Grab Cut algorithm, put forwards a interactive image segmentation algorithm, making full use of the user’s subjective goals.3. Put forward the SIFT local feature and color histogram as the mixed feature extraction scheme, extracting and matching clothing commodities image feature, testing the algorithm as modules.4. Refer to the Android programming principle, grasp the basic APIs and use of SQLite database. Implement the system as modules, in order to achieve a better user experience, UI interface design is optimized further.Finally, this system is tested and runs stably, basically satisfies the demand.The main modules of the clothing retrieval system is account management module,take photos module, image processing module and result display module. Finally, The system successfully provides the users with the convenient modern mobile shopping experience.