Magdas Dana Alina, Hategan Ariana Raluca, David Maria, Berghian-Grosan Camelia
National Institute for Research and Development of Isotopic and Molecular Technologies, 67-103 Donat Street, 400293 Cluj-Napoca, Romania.
Faculty of Physics, Babeș-Bolyai University, Kogălniceanu 1, 400084 Cluj-Napoca, Romania.
Foods. 2025 May 19;14(10):1808. doi: 10.3390/foods14101808.
Artificial intelligence (AI) tends to be extensively used to develop reliable, fast, and inexpensive tools for authenticity control. Initially applied for food differentiation as an alternative to statistical methods, AI tools opened a new dimension in adulteration identification based on images. This comprehensive review aims to emphasize the main pillars for applying AI for food authentication: (i) food classification; (ii) detection of subtle adulteration through extraneous ingredient addition/substitution; and (iii) fast recognition tools development based on image processing. As opposed to statistical methods, AI proves to be a valuable tool for quality and authenticity assessment, especially for input data represented by digital images. This review highlights the successful application of AI on data obtained through laborious, highly sensitive analytical methods up to very easy-to-record data by non-experimented personnel (i.e., image acquisition). The enhanced capability of AI can substitute the need for expensive and time-consuming analysis to generate the same conclusion.
人工智能（AI）倾向于被广泛用于开发可靠、快速且廉价的真实性控制工具。人工智能工具最初作为统计方法的替代方法应用于食品鉴别，开启了基于图像的掺假识别新领域。这篇综述旨在强调将人工智能应用于食品认证的主要支柱：（i）食品分类；（ii）通过添加/替代外来成分检测细微掺假；以及（iii）基于图像处理开发快速识别工具。与统计方法不同，人工智能被证明是质量和真实性评估的宝贵工具，特别是对于以数字图像为代表的输入数据。这篇综述突出了人工智能在通过费力、高灵敏度分析方法获得的数据上的成功应用，直至非专业人员非常容易记录的数据（即图像采集）。人工智能增强的能力可以替代昂贵且耗时的分析来得出相同结论。
