Online Virtual Try-On using Computer Vision

Online clothing shopping has grown in popularity among millions of people worldwide in recent years, particularly during the COVID-19 pandemic. Online shopping still has some room for improvement compared to the in-store shopping experience. Some people are still hesitant to buy clothes online because they’re not sure how they look and fit on them. Online shopping websites have models to show their product, but there is no way to check how they actually look on customers. Customers who shop online tend to return the item if they do not like it after trying it at home. The company’s resources are wasted by providing free returns to their customers. Hence, virtual try-on software is a desperate need for both the online shopping customer and e-Commerce clothing businesses. We have built an application using Flask, OpenCV and Mediapipe. The software works with gadgets that have a web browser, Internet access, and a camera that is built-in or attached. The client can select the apparel to test on virtually and then the software will map it to the user’s body using openCV and MediaPipe. The user can try clothes virtually, review them, and decide whether to buy it or not. This eliminates the crucial problem of not being able to try on clothes while shopping online. Our software provides a quick, simple and accurate method of trying on clothing as a result, and we think it will be crucial for the success of the online industry.