A review of three-dimensional vision techniques in food and agriculture applications

Abstract

In recent years, three-dimensional (3D) machine vision techniques have been widely employed in agriculture and food systems, leveraging advanced deep learning technologies. However, with the rapid development of three-dimensional (3D) imaging techniques, the lack of a systematic review has hindered our ability to identify the most suitable imaging systems for specific agricultural and food applications. In this review, a variety of 3D imaging techniques are introduced, with their working principles and applications in agriculture and food systems. These techniques include Structure lighting-based 3D imaging, Multiview 3D imaging system, Time of Flight (ToF)-based 3D imaging system, Lighting Detection and Ranging (LiDAR), and Depth estimation from monocular image. Furthermore, the three-dimensional image analysis methods applied to these 3D imaging techniques are described and discussed in this review.

Keywords

• 3D imaging
• RGB-D imaging
• Stereo imaging
• Deep learning
• Point cloud analysis