A Virtual Trial Room using Pose Estimation and Homography

Abstract

Rapid urbanization and digitization, increasing income and lifestyle changes of the middle class are leading to a major revolution in the retail industry, especially shopping malls. Today almost every city in India has several shopping malls and with the increase in population, the queues for trying on clothes in these shopping malls are increasing rapidly. This paper is a research in overcoming this problem and enhancing the customer's shopping experience. We have built an android based mobile application along with the technologies of OpenCV and TensorFlow lite which will allow the customers to try on the clothes virtually without waiting in a queue for a long time. The application uses OpenCV to map the clothes onto the customer's body, the customer can then have a look at it and decide whether to buy the clothes or not. The results obtained show that the mapping is done accurately onto the user's body and that it can avoid the need to go to trial rooms. Thus our application provides a quick, easy, and accurate way to try clothes and thus we believe it will prove instrumental in shaping the retail industry.