Automated Super Shop using image processing (Python)

Abstract

Unstaffed retail shop is been emerging out in the past years and has significantly affected conventional shopping styles. In this area, unmanned retail container plays an very important role, it can highly influence the shopping experience of user. the traditional way based on weighing sensors can’t sense what the customer is taking. This paper proposes a smart unstaffed retail shop scheme based on Image processing using python, aiming at exploring the feasibility of implementing the unstaffed retail shopping. Based on the data set of images in different scenarios that  includes different types of stock keeping unit (SKU) with variable sizes, an end-to-end classification model of unstaffed shop  trained by the method is developed for SKU recognition & counting, and the proposed solution in this study is able to achieve counting and recognition accuracy on the test data table, which indicates that the system can make up a good choice over deficiency of traditional unmanned container.