Recent Developments of Machine Learning Combined 3D Reconstruction in Agricultural and Food Sectors

The integration of machine learning (ML) with 3D reconstruction can enable the precise and intelligent monitoring of management strategies across the agri-food sector more efficiently. The capability to generate accurate 3D virtual models has become essential for optimizing agri-food production towards automation. ML techniques play a critical role in enhancing the accuracy of 3D reconstruction by improving feature extraction, segmentation, and real-time analysis of data acquired in various agri-food production environments. This survey examines recent developments in ML-driven 3D reconstruction within the agricultural and food sectors, focusing on the application of ML for both reconstructing high-fidelity 3D models from imaging data and utilizing these reconstructed models for data-driven decision-making in agri-food production. A comprehensive analysis of the advanced and emerging techniques for ML-based 3D reconstruction in precision agriculture, livestock farming, and food manufacturing over the past five years is provided. Finally, the challenges and potential future trends of 3D reconstruction in agriculture and food sectors are discussed.