This article takes the mural “Playing Music Painting” in the tomb chamber of Yan Defei from the Tang Dynasty as its subject. It systematically interprets the morphological characteristics and cultural connotations of the costumes worn by female musicians in the painting. Based on this, digital restoration methods are introduced. By combining image analysis, clothing history research, and CLO 3D modeling technology, a structured digital restoration of women’s geisha costumes is conducted to explore the value of digital technology for ancient clothing research and cultural heritage protection.
The growing demand for personalised apparel, combined with advances in artificial intelligence (AI), is reshaping the apparel customisation industry chain. However, existing studies mostly focus on the isolated optimisation of individual stages (e.g., design, production, or marketing) and lack an integrated perspective on multi-agent collaboration across the entire value chain. To address this gap, this paper presents a systematic review conducted in accordance with PRISMA guidelines. From the Web of Science, Scopus, IEEE Xplore, and CNKI databases, 35 eligible studies were selected and critically synthesised. The review identifies five core interconnected challenges: high demand variability, complex product/bill scheduling with frequent disruptions, fragmented information across stages, inefficient material utilisation, and slow design-to-production lead times. Key research gaps are also identified, including the integration of sustainability metrics, adaptation for small and medium-sized enterprises, cross-domain fashion trend integration, and human-AI collaboration mechanisms. Based on these synthesised findings, a conceptual five-layer AI-agent collaborative architecture is discussed as a potential future research direction to achieve end-to-end coordination across design, sourcing, production, inventory, and marketing. This conceptual framework synthesises current technological trends but requires empirical validation in real-world settings.
Occupational exposure to solar heat radiation is a major contributor to heat stress among outdoor workers, particularly in agriculture, where physically demanding tasks are performed under high ambient temperatures and direct sunlight. Infrared (IR) radiation represents a significant external heat load that can impair thermoregulation even when convective and evaporative cooling are present. A key research gap addressed in this paper is the limited understanding of how garment materials simultaneously influence both infrared (IR) transmission and IR absorption, and how these competing mechanisms affect net heat stress. This study evaluated garment materials and a prototype protective vest designed to reduce heat stress from solar IR exposure. Phase I measured IR penetration and IR absorption for cotton, polyester, paper, and Mylar using controlled radiant heat exposure and thermocouple measurements. Results showed that conventional fabrics reduced IR transmission but absorbed substantial radiant energy, leading to net heat gains of approximately 39–40%. In contrast, Mylar reduced IR penetration by 50% and offered a net heat loss of 15%. Phase II integrated the optimal material into a prototype vest incorporating the reflective Mylar foil and an 8.5 mm ventilation gap, which was evaluated using an inflatable thermal mannequin. Under IR exposure, the results showed that the prototype vest reduced the mannequin’s skin temperature by 26.4% and total body heat gain by 63.9% compared with a nude control and a standard polyester vest. These findings demonstrate that separating IR attenuation from absorption is critical in garment design. The results support the use of reflective, ventilated vest systems as a strategy for reducing occupational solar heat exposure.
3D dressed-modelling has become a core technology in fashion design, digital human modelling, and virtual try-on applications. Existing image-based multi-view garment modelling methods primarily emphasise visual realism while often overlooking dimensional accuracy. To address this issue, this study proposes a high-precision 3D garment reconstruction framework based on multi-view images of dressed subjects. A Mask-CNN neural network is utilised for instance segmentation of clothed human images, enabling the extraction of 2D garment-only regions. The model achieves a low training loss of 0.121 3, indicating reliable segmentation performance. Subsequently, a multi-view reconstruction approach is employed to generate the 3D garment model, followed by Poisson surface reconstruction to enhance geometric smoothness and realism. Comparative analysis between the reconstructed 3D garment and the actual physical garment demonstrates a waist circumference error of only 2.01%, confirming the high geometric accuracy and practical applicability of the proposed method. Unlike reconstruction pipelines that primarily target visual plausibility, the present work explicitly evaluates whether the reconstructed dressed-garment geometry preserves physically meaningful garment dimensions. This metric-oriented perspective is important for subsequent use in digital garment engineering and pattern-related applications.
The carding process plays a key role in determining sliver quality in fibrous material production. However, the dynamic aspects of the process, particularly the influence of time delays and internal positive feedback loops on the achievement of the needed sliver, remain insufficiently studied, which hinders the development of effective control systems. To address this limitation, the aim of this work is to develop a theoretical mathematical model that establishes a quantitative relationship between design and operational parameters (such as doffing ratio and draft) and unevenness indicators. The study examines a dynamic model of fibrous mass motion that accounts for the interaction between the main cylinder and the doffer cylinder using methods from control theory to derive transfer functions. The analysis reveals that the system dynamics, described by a delay equation, exhibit stepwise exponential transient responses where the settling time is dependent on the doffing ratio. Furthermore, frequency analysis demonstrates that machine-selected leveling properties, the maximum leveling effect is achieved for unevenness wavelengths equal to half the cylinder circumference or a multiple thereof, while low-frequency (long-wave) variations are practically uncorrected. These findings allow for a quantitative assessment of the influence of key parameters on sliver unevenness and provide a foundation to optimize process settings and develop automatic control systems that compensate for undesirable dynamic effects.
