@article{ZHANG2026, 
author = {Shengmao ZHANG and Xinyu AI and Fei WANG and Tianfei CHENG and Quanyou GUO and Wei FAN},
title = {Advances in the application of hyperspectral imaging for nondestructive fish quality assessment},
year = {2026},
journal = {Transactions of the Chinese Society of Agricultural Engineering},
volume = {42},
number = {9},
pages = {383-396},
keywords = {hyperspectral, fish quality, nondestructive testing, characteristic wavelengths, aquaculture},
url = {https://www.sciopen.com/article/10.11975/j.issn.1002-6819.202509157},
doi = {10.11975/j.issn.1002-6819.202509157},
abstract = {Aquatic products are often required to assess their quality and safety during fish processing. However, conventional chemical assays and sensory evaluation cannot fully meet the ever-increasing needs in modern aquaculture. Among them, chemical analysis is usually destructive, labor-intensive, and low-speed for rapid or large-scale inspection. While the sensory evaluation can depend heavily on subjective experience and judgment, leading to its reliability in quality assessment. Alternatively, hyperspectral imaging (HSI) has emerged as an attractive nondestructive approach to evaluate fish quality. The imaging can be expected to integrate with spectroscopy, thereby simultaneously capturing spatial and spectral information from fish muscle. The visible, near infrared, and short-wave infrared regions can be covered over the broad range of quality variations in fish tissues. Physicochemical changes include the freshness, moisture distribution, lipid characteristics, and microbial spoilage. Spectral signals can provide the spatial distribution to evaluate multiple quality attributes without damage to the sample. The HSI technique can also share the superiority to conventional imaging and destructive analytical methods in fish-quality detection. In this review, the recent progress was systematically summarized in the application of HSI into fish freshness, moisture, fat, microbial spoilage, and parasite contamination. Much emphasis was also placed on the spectral ranges most frequently in previous research. The approaches were selected based on the informative wavelengths, modeling, and their predictive performance. Meanwhile, some attention was also given to the main barriers to the broader use of HSI. Among them, it was lacking in consistency in spectral acquisition and instrument standards, data processing and interpretation, adaptability to different sample-handling conditions and real application scenarios, multi-index fusion and model generalization, equipment costs, and unified standards. Taken together, it was also required to translate laboratory findings into industrial practice. Looking ahead, future advances were likely to involve closer equipment integration, higher levels of intelligence, multimodal data fusion, closed-loop process control, and deeper application in fish-quality evaluation. At the same time, the standardized technical frameworks and more complete industrial systems are also essential for wider application. Furthermore, the HSI can be integrated with the sensing technologies and intelligent decision-making tools for high detection efficiency, robustness, and practical applicability. Overall, this finding can offer a useful reference to advance the intelligent upgrading of fish-product processing for the high quality and safety of aquatic products.}
}