3D garment virtual fitting technology is a computer technology that can convert 2D garment patterns into 3D models. This technology has wide applications in the garment industry due to its advantages of high efficiency, low cost, and high simulation accuracy. Enterprises or research institutions can use it to quickly obtain garment fitting effects. Additionally, the performance of garments can be predicted and evaluated through this technology, thus greatly reducing the cost of garment development.
The 3D virtual fitting technology composition mainly includes three parts: 3D body measurement technology, 3D body modeling technology, and 3D garment modeling technology. These three parts together determine the final simulation effect. Data driving, machine learning, and other technologies enable computers to identify, segment, and process the data obtained from body measurements more efficiently. At the same time, many parametric human models and clothing generation models suitable for interactive scenarios have emerged.
The structure design, print design, and customercentered modular collaborative design method Based on 3D virtual fitting technology make the garment design and development progress more efficient. Using 3D virtual fitting technology to invert the 3D model of garments not only makes it possible to obtain 2D patterns of complex threedimensional shapes, but also precisely locate and divide prints on the patterns. The customercentered modular collaborative design mode employs modularized design. This mode provides a communication platform for customers, designers, and evaluation experts so that they can collaborate to complete the design and obtain a product that satisfies customers.
For apparel aesthetics evaluation, by building a virtual garment drape test platform based on 3D virtual fitting technology, a rapid evaluation of garment drapability can be achieved. In the garment fit evaluation area, studies have used the key pressure points of virtual garments and body dimensions obtained through 3D virtual fitting technology as input data to train a neural networkbased garment fit prediction model. The prediction accuracy is influenced by the type and scale of the input parameters and model algorithm. For garment comfort evaluation, studies mainly focus on pressure comfort evaluation, and its accuracy is influenced by factors such as the sampling method, number of key points, and state of wearers. In addition, the air layer distribution under the garment obtained through 3D virtual fitting technology can also be used to evaluate the thermal and moisture comfort of garments, but the overall accuracy is not high enough.
In the postepidemic era, 3D virtual fitting technology faces corresponding challenges while having broad application prospects. In the future, it is necessary to standardize and unify the parameters of virtual fabrics. When evaluating garments, researchers should build virtual models with human soft tissue characteristics in specific scenarios and take the physical characteristics of garment parts into account. In addition, through the introduction of artificial intelligence, deep learning, and other technologies, it is possible to optimize and realize automatic sampling. Then, we build a whole performance evaluation model for virtual garments to improve evaluation accuracy and efficiency.