Visualization and storage algorithms associated with processing point cloud data
First Claim
1. A method for organizing tuples of a point cloud wherein said tuples contain data related to physical features comprising:
- constructing a virtual grid having a plurality of intersections over said point cloud of tuples, said intersections spaced in accordance with a desired spatial frequency, each intersection of said intersections associated with a predetermined data structure,subdividing said virtual grid into tiles of a predetermined size, each tile of said tiles containing a plurality of said intersections,developing a plurality of separate tile records, one tile record for each said tile,storing tuple data from said tuples falling in each said tile in a respective said tile record so said tuple data related to a selected region of said point cloud is quickly retrievable by accessing only said tuple data stored in those said tile records associated with said selected region,defining an overlap tile along each boundary where two adjacent said tiles meet, said overlap tile extending into each of said adjacent tiles so that said tuples in said overlap tile may be treated as belonging to one or the other or both of said adjacent tiles,when a region in said point cloud is selected, triangulating only said tuples in said tiles and said overlap tiles of a selected said region into a linked triangulated network, with tuples in said overlap tiles along edges of said selected region allowing completion of a triangulation process, eliminating a need for triangulating all tuples in an adjacent tile and increasing computational speed of the process,populating at least most of said predetermined data structures with data interpolated from said triangulated network, said triangulated network extending into said overlap tiles, and,using said data in said data structures to develop a display of said physical features in said selected region.
0 Assignments
0 Petitions
Accused Products
Abstract
A plurality of algorithms for interpreting and displaying point cloud data are disclosed. Typically, a virtual grid is constructed over point cloud data, with intersections of the grid at a predetermined spacing that may be dependent on density of the point cloud. Several methods are implemented that utilize tuple data from tuples near a respective intersection to populate intersections of the grid with data. In most embodiments, the intersections of the grid correspond to pixels in a raster image. The tuple data may be manipulated to determine Nyquist spacing of points in a point cloud and portions thereof, develop monoscopic and stereoscopic images, and enhance selected features of an image. Additional algorithms are disclosed that reduce storage and computational requirements during processing of point clouds. Such algorithms use a tiled grid structure wherein processing and data storage proceeds in one or a few of the tiles at a time.
86 Citations
22 Claims
-
1. A method for organizing tuples of a point cloud wherein said tuples contain data related to physical features comprising:
-
constructing a virtual grid having a plurality of intersections over said point cloud of tuples, said intersections spaced in accordance with a desired spatial frequency, each intersection of said intersections associated with a predetermined data structure, subdividing said virtual grid into tiles of a predetermined size, each tile of said tiles containing a plurality of said intersections, developing a plurality of separate tile records, one tile record for each said tile, storing tuple data from said tuples falling in each said tile in a respective said tile record so said tuple data related to a selected region of said point cloud is quickly retrievable by accessing only said tuple data stored in those said tile records associated with said selected region, defining an overlap tile along each boundary where two adjacent said tiles meet, said overlap tile extending into each of said adjacent tiles so that said tuples in said overlap tile may be treated as belonging to one or the other or both of said adjacent tiles, when a region in said point cloud is selected, triangulating only said tuples in said tiles and said overlap tiles of a selected said region into a linked triangulated network, with tuples in said overlap tiles along edges of said selected region allowing completion of a triangulation process, eliminating a need for triangulating all tuples in an adjacent tile and increasing computational speed of the process, populating at least most of said predetermined data structures with data interpolated from said triangulated network, said triangulated network extending into said overlap tiles, and, using said data in said data structures to develop a display of said physical features in said selected region. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22)
-
Specification