Research On The Usage Of Ladar And Multisensor Fusion Technology

Autonomous Land Vehicle (ALV) is a kind of intelligent mobile robot, which can be driven on road or cross-country environment autonomously and continuously. ALV and its corresponding technology is of great value in both research and application area. This thesis aims at the research on the usage of ladar and multisensor information fusion technology applied in ALV.At first, the geometric model of ladar is studied. The geometric model of multi-layer ladar is built in the background of LD ML. The transformation formula from range data to 3-D coordination system has been deduced. A new adaptable and restricted method of clustering analysis, named as Adaptable Restricted Clustering Algorithm (ARCA), is presented upon the discussion of single-frame range data processing. The detection method of roads and obstacles based on the result of ARCA is then proposed.Considering at the disadvantages caused by single-frame range data processing, the methods of building a grid map with multi-frame range data and obstacles detection algorithm emerged from grid map have been studied and realized. The discussion about the method of local road boundary detection via Hough-transform is followed.Finally, the fusion means of ladar and vision sensors is presented. The fusion happens in both the data and feature levels. Also we give a novel way of importing meaningful color feature to the grid map built from ladar data.