Computer Vision Across the Garment Lifecycle: Applications and Sustainability Implications
The garment industry faces sustainability challenges involving resource consumption, material waste, product returns, and textile disposal. This structured narrative review examines CV (computer vision) applications across four garment lifecycle stages: design and sample development, manufacturing and quality control, fashion retail and consumption, and post-consumer reuse and recycling. The literature is organized into eight CV task categories and six algorithm families using independent, multi-label coding. Sustainability claims are evaluated by distinguishing empirical lifecycle assessments, factory measurements, consumer behaviour evidence, model-based predictions, and qualitative inference. Comparative studies suggest that digital sampling can reduce environmental burdens, although these outcomes cannot be attributed to CV alone. Manufacturing research reports improved inspection performance and selected operational benefits. In contrast, retail and post-consumer studies predominantly emphasize visual quality, classification accuracy, and processing efficiency, with limited evidence of reduced returns, increased material recovery, or lower emissions. The review identifies a persistent gap between technical performance and verified sustainability outcomes, alongside barriers involving cross-domain robustness, implementation costs, and data interoperability. Future research should integrate CV performance metrics with field measurements and lifecycle assessment to evaluate net environmental impacts.
Authors
- Rongrong Cui (ORCID: https://orcid.org/0009-0002-0016-8631)
- Xiaowan Sun (ORCID: https://orcid.org/0000-0003-1312-0901)
Institutions
- Zhejiang Sci-Tech University (CN)
- Zhejiang Institute of Modern Textile Industry (CN)
- Keyi College of Zhejiang Sci-Tech University (CN)
Publication Details
- Journal
- Sustainability
- Published
- 2026-11-16
- DOI
- https://doi.org/10.3390/su182211639
- Primary Topic
- Cultural and Historical Studies
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Science and Technology Department of Zhejiang Province