Overview of image-based 3D vision systems for agricultural applications

Various agricultural robots exist which are intended to help farmers to make decisions. Vision systems are an integral and essential part of these agricultural robots to perceive and evaluate the work space, inspect crops and detect objects. The use of 3D vision systems can provide a 3D reconstructed model which will supply detailed information including depth estimation about a crop or an object or environmental structures. However, these vision systems have some drawbacks which are limiting their use in an agricultural area. The limiting factors are dealing with lighting conditions, navigating unstructured environment, harvesting speed, occlusion and overlapping of fruits, accuracy of the system etc. This paper investigates the limitations and challenges of the state-of-the-art of 3D image reconstruction techniques in agricultural applications. Active and passive approaches have been reviewed. Factors to be considered while designing the vision system have been identified.