Self-contained mapping and positioning system utilizing point cloud data

Abstract

A self-contained mapping and positioning system for underground mining is provided that is capable of mapping the topography of a region, such as a mine tunnel, and further being able to use the mapped data to determine the position of an object, such as a mining vehicle, within the mine tunnel. The system includes an inertial navigation system, a central processing unit, a three-dimensional database, a three-dimensional camera system, an operator console and a survey system having a three-dimensional laser scanner. The survey system using the three-dimensional laser scanner produces point cloud data, i.e., a set of data points representing the topography of the region. The point cloud data is stored within a storage device until the entire region is mapped and then transmitted to the operator console to be post processed. After post processing, the data is exported to the three-dimensional database and the indexed for ease of use by the central processing unit. To determine the position of the object within the region, the system of the instant invention initializes the object in its current position. The object is then either remotely or directly guided to another position from the current position where it is brought to an estimated position by the inertial navigation system or is remotely controlled for performing work, e.g., drilling a mine heading. After a predetermined time of moving the object, there is an error in the position of the object and the operator console makes a call to the central processing unit on board the object to automatically calculate the true position based on the point cloud data stored within the database to update the position of the object. This is done by approximating a search range for the database according to the estimated position. A subset of data corresponding to the search range is removed from the database. The three-dimensional camera system then scans images of the surface in proximity to the object and converts the images to a patch of point cloud data. The patch of point cloud data is then matched against the subset of data corresponding to the search range of point cloud data removed from the three-dimensional database until there is less than a predetermined minimum error distance. At that point, the true position of the object is known. This new position is then put back into the inertial navigation system and the positional data of the object is updated with the correct positional data.