Digital image-based chemometrics for food analysis: a practical tutorial and roadmap

Abstract

Digital images have become a powerful tool for developing analytical methods in food quality control. Unlike conventional analytical signals, images can be processed to extract relevant chemical information, with chemometric techniques enhancing their utility. This review synthesizes applications of digital imaging in food analysis, providing a roadmap from univariate methods to multivariate classification/calibration approaches, illustrated through three case studies demonstrating their potential for food safety and quality. However, the field faces critical challenges, particularly the lack of methodological standardization, as evidenced by diverse applications in literature. Addressing this gap is essential to ensure reliability and reproducibility. Furthermore, the review highlights recent advances, such as hybrid color descriptors, chromaticity maps, deep learning architectures, and time-resolved RGB imaging, that improve the robustness and applicability of these techniques in food science.