Non-contact evaluation of liquid transport behavior in sportswear textiles: a SAM-enhanced infrared thermography approach

Purpose This study establishes a non-contact infrared (IR) thermography framework to overcome color and/or texture biases plaguing traditional textile liquid transport evaluation. By integrating Segment Anything Model (SAM)-based image segmentation with thermal dynamics analysis, we aim to (1) develop optical-property-independent quantification of directional capillary transport; (2) identify critical performance determinants through temporal-spatial feature extraction and (3) create standardized metrics for next-generation sportswear optimization. The methodology enables real-time, interference-free assessment of synthetic fabric liquid management, addressing industry needs for objective performance benchmarking. Ultimately, this advances data-driven design of functional textiles with engineered capillary networks, moving beyond empirical trial-and-error approaches in perspiration management system development. Design/methodology/approach A non-contact IR thermography system integrated with SAM was developed to overcome color/texture biases in traditional textile evaluation. IR imaging captured real-time thermal dynamics during simulated sweat transport (10 fabric specimens), while the SAM algorithm segmented wetting fronts independent of optical properties. Six capillary metrics (Wetting Time (WT), Maximum Wetting Area (MWA), Maximum Wetting Rate, etc.) were extracted from temporal-spatial thermal profiles. Principal component analysis (PCA) reduced dimensionality to two critical factors (WT: temporal efficiency; MWA: spatial capacity), explaining 82.74% variance. Validation included heatmap visualization and comparative benchmarking against gravimetric methods. The framework enables standardized, real-time characterization of directional liquid transport without contact interference. Findings The SAM-enhanced IR thermography system successfully quantified directional capillary transport non-invasively, overcoming color/texture biases of conventional methods. PCA identified two critical dimensions: WT (63.21%) for temporal efficiency and MWA (19.53%) for spatial absorption capacity, jointly explaining 82.74% of variance. Heatmap analysis validated these dominant factors, revealing synthetic blends (e.g. Coolmax/polypropylene) outperformed natural fiber composites in transport efficiency. The framework enabled real-time monitoring without contact interference, demonstrating superior reliability compared to gravimetric techniques. Results provide a physics-based metric system for optimizing pore structures and hydrophilic-hydrophobic fiber arrangements, advancing standardized evaluation of liquid management in functional sportswear textiles. Research limitations/implications While demonstrating efficacy on ten synthetic sportswear fabrics, the method requires validation across broader textile classes (e.g. natural fiber blends) and environmental conditions (humidity/temperature variations). Current metrics focus on unidirectional transport; multi-axial capillary dynamics remain uncharacterized. The SAM-IR framework establishes a standardized evaluation protocol for textile engineers to objectively optimize liquid management via pore geometry tuning and fiber hydrophilicity gradients. This enables data-driven development of advanced perspiration-wicking fabrics, reducing reliance on empirical trial-and-error approaches. Future integration with biomechanical models could personalize sportswear design based on athlete-specific thermoregulatory demands. Practical implications This non-contact framework enables textile manufacturers to objectively benchmark liquid transport performance across diverse fabrics (dark/patterned textiles included), overcoming traditional color/texture limitations. The established metrics (WT/MWA) guide data-driven optimization of pore architectures and hydrophilic-hydrophobic fiber ratios for enhanced perspiration management. Quality control labs can implement this real-time system for rapid sportswear grading (e.g. LTEI certification), reducing evaluation time by 68% versus gravimetric methods. Sportswear brands may leverage heatmap-derived capillary activation patterns to engineer zone-specific moisture routing in compression garments, improving athlete thermoregulation. The methodology paves the way for smart textile development with embedded IR sensors for on-body liquid transport monitoring. Social implications This technology promotes equitable sportswear development by eliminating color-based evaluation biases, ensuring athletes wearing dark-colored apparel receive objectively assessed moisture-wicking gear. Enhanced thermoregulation fabrics reduce heat stress risks during intense exercise, particularly benefiting youth sports participants susceptible to overheating. The non-contact method facilitates customized moisture management designs for adaptive athletes using prosthetics. Environmentally, optimized liquid transport extends garment lifespan, reducing microfiber pollution from frequent washing. Consumers gain access to scientifically validated performance data, empowering informed purchases for medical needs like eczema management. By shifting textile evaluation from empirical to physics-based metrics, this framework advances inclusive, sustainable innovation in functional apparel engineering. Originality/value This study pioneers the integration of SAM with IR thermography to establish the first non-contact, color-agnostic framework for quantifying directional liquid transport in textiles. The novel SAM-enabled wetting front segmentation overcomes optical property limitations of traditional methods, while real-time thermal dynamics capture reveals critical temporal-spatial transport mechanisms. By defining two physics-based principal components (WT: temporal efficiency; MWA: spatial capacity), the work provides a paradigm shift from empirical textile evaluation to data-driven design. This methodology sets new standards for functional sportswear development, enabling precise optimization of capillary networks and pore geometries without contact interference – a breakthrough in perspiration management research.