The rapid expansion of e-commerce platforms and digital fashion marketplaces has significantly increased the demand for intelligent and interactive shopping solutions. Traditional online clothing stores rely on static images and size charts that often fail to provide customers with a realistic understanding of how outfits will appear on their bodies. This limitation frequently leads to uncertainty during purchasing decisions and contributes to high product return rates in the fashion industry. To address these challenges, this paper presents an AI and computer vision-based framework for real-time virtual outfit visualization that enables users to digitally try on clothes using a webcam without physically wearing them. The proposed system integrates artificial intelligence, computer vision, and image processing techniques to create a real-time virtual try-on experience. The framework utilizes human pose estimation, body segmentation, and outfit overlay algorithms to accurately map digital clothing items onto the user’s body. The system continuously tracks body movements and dynamically adjusts outfit alignment according to posture and orientation changes. OpenCV and MediaPipe Pose are employed for real-time body detection and tracking, while image transformation techniques are used for clothing alignment and rendering. The developed application operates within a lightweight web-based environment using the Flask framework, allowing accessibility through standard computing devices without specialized hardware requirements. Experimental analysis demonstrates that the proposed system achieves stable visualization performance with low latency and accurate outfit positioning under various environmental conditions. The framework improves user engagement in online shopping environments and provides a scalable solution for virtual fashion applications. The proposed approach also has potential applications in augmented reality shopping systems, digital fashion retailing, and personalized clothing recommendation platforms.
S et al. (Sat,) studied this question.