Key points are not available for this paper at this time.
Change detection has been increasingly studied in remote and close-range sensing in the last decades, driven by its importance in environment monitoring and database updating. Due to the development of sensing technologies, data acquisition has become more accessible and affordable and thus more data from various sensing platforms have become available. Thanks to structure-from-motion photogrammetry and lidar technologies, 3D change detection from point cloud data is drawing considerable attention in recent years. Motivated by the lack of a comprehensive review of 3D change detection in the urban environment, this paper reviews the latest developments in urban object change detection using point cloud data. In particular, four types of objects, namely building, street scene, urban tree, and construction site, are analysed in-depth. The use of different data sources for each object-of-interest and the open-source data with change labels are summarised. Then the change detection methods are thoroughly reviewed at pixel, point, voxel, segment and object levels, whose pros and cons are analysed in detail. Moreover, the challenges and opportunities brought by point cloud data and new methods, such as Siamese network deep learning, are discussed for future considerations.
Xiao et al. (Thu,) studied this question.