An Application Domain-Based Review of Proposed Approaches Based on Lidar Image Registration

Light detecting and Ranging (LiDAR) technology has lately emerged as one of the most innovative fields in laser scanning, object detecting, remote sensing systems. The technique has proven to be attractive since it has surface penetrative abilities which enables it to locate structures or areas of interest down to the level of millimeter. The registration of a three-dimensional (3D) point cloud has proven to be an important step in many 3D modeling and mapping applications. It can also show differences and abnormalities like surface degradation and vegetation development, something that has led to its varied applications in the fields of agriculture, road safety patterns, archeology, etc. Since the scene complexity, data source, and application of existing approaches varied greatly, contemporary practices in distinct point cloud registration tasks remain ad hoc. Recent improvements in deep learning-based approaches have shown promising results in evaluating stiffness and similarity. In this paper, several applications of lidar in various fields have been presented with a broad comparison among different calibration practices.