Application of virtual reality in e-commerce: Taking the experience of trying on sports equipment as an example
Lin Gan · Molecular & cellular biomechanics · 2025
Traditional e-commerce platforms have the problem that users cannot try on sports equipment in person, and it is difficult for consumers to perceive its size, comfort and dynamic performance before purchasing. This limitation leads to high return rates and difficult purchasing decisions. This paper introduces a virtual try-on solution with higher accuracy and more immersion. After using 3D scanning technology to obtain the user’s body data and combining it with SMPL (Skinned Multi-Person Linear Model) to generate the user’s body model, a posture optimization algorithm is used to adjust the dynamic posture of the user model and the PoseNet optimization model is used to adapt it to the user’s dynamic motion scenes. Next, Unity Physics is used to achieve the dynamic performance of sports equipment materials, high-definition texture mapping technology is used to reproduce the visual effects of equipment materials to ensure that the appearance is consistent with reality, and sports scenes are constructed to simulate the actual performance of equipment in different environments. Users can use motion capture devices to simulate running, jumping and other movements to feel the suitability of sports equipment. Then, based on the user’s body shape data and sports scene preferences, Deep Q-Learning is used to recommend sports equipment options suitable for the user. Finally, the system adjusts the virtual try-on experience in real time, showing a variety of combination effects, helping users quickly find their favorite products. Experiments show that the performance error between virtual equipment and real equipment is only 1.35%, the virtual try-on pass rate exceeds 90%, and the return rate is less than 10%, which verifies the feasibility of virtual reality technology in e-commerce and improves users’ online shopping experience.