Advances in 2D and 3D machine vision technologies for morphological characterization of granular food products: from laboratory to application

Granular food products, as essential components of global agricultural systems, require precise morphological evaluation (e.g., color, size, and texture) for quality grading and safety assurance. Traditional manual inspection is inefficient and subjective in industrial-scale operations. This review examines the advancements in 2D and 3D machine vision technologies for morphological quality assessment. Two-dimensional methods enable efficient surface analysis through image processing, achieving over 90% accuracy in detecting color anomalies, geometric parameters, and defects (e.g., damage, mildew, adhesion), but lack spatial depth. Conversely, 3D reconstruction via point cloud analysis delivers precise volumetric measurements (thickness and volume) and fine texture characterization, yet faces challenges in cost optimization and computational efficiency. The complementary strengths of both technologies are evident: 2D vision excels in rapid surface screening, whereas 3D techniques resolve complex structural details. Emerging integration strategies, including digital twins and digital fingerprinting, show promise for quality monitoring, spanning from surface attributes to internal features. By synergizing the cost-effectiveness of 2D systems with 3D spatial precision, future systems could enable end-to-end quality control across production chains. This review compares technological capabilities, outlines optimization paths, and proposes an intelligent inspection framework, providing theoretical support for automated quality management in granular food industries.