Method and apparatus for automatically generating a site model
AssigneeSarnoff Corporation
InventorsYanlin Guo, Rakesh Kumar, Stephen Charles Hsu, Harpreet Singh Sawhney
Life of this patent
Estimating the term…
Abstract
A method and apparatus for automatically combining aerial images and oblique images to form a three-dimensional (3D) site model. The apparatus or method is supplied with aerial and oblique imagery. The imagery is processed to identify building boundaries and outlines as well as to produce a depth map. The building boundaries and the depth map may be combined to form a 3D plan view model or used separately as a 2D plan view model. The imagery and plan view model is further processed to determine roof models for the buildings in the scene. The result is a 3D site model having buildings represented rectangular boxes with accurately defined roof shapes.
Claims26
A method tot generating a three-dimensional site model of a site that contains objects comprising:
- (a) providing at least one of aerial imagery and oblique imagery of the site;
- (b) processing at least one of the aerial imagery and the oblique imagery to generate a plan view model comprising object outlines,
- wherein said processing further comprises producing an elevation map using key video images;
- analyzing the elevation map to identify surface patches;
- computing local orientation of the surface patches within the elevation map, aggregating the surface patches for the key video images with respect to the local orientation of the surface patches;
- aggregating elevation information amongst the key video images to produce a three-dimensional structure;
- processing the three-dimensional structure into a plan view model;
- and
- (c) producing from the plan view model a three-dimensional site model comprising blocks that represent objects within the site,
- wherein the objects within the site include at least one building, the rectilinear outline of the at least one building being determined by:
- computing an orientation histogram for the initial rectangle representing the building and selecting the angle of orientation with the strongest peak as the dominant orientation of the building;
- computing a gradient field based on image intensity of the building;
- integrating the gradient field along and perpendicular to the dominant orientation of the building;
- fitting straight lines to local maximums of the gradient field along a each direction;
- detecting knot points at the crossings of the straight lines;
- dividing the initial rectangle into a plurality of smaller rectangles based on the knot points;
- and
- traversing the smaller rectangles to find the rectilinear outline of the building.
The method if claim wherein the objects in the site are buildings and the method further comprises:
- (d) processing the oblique imagery to produce a roof model for each building in the site.
The method of clam 1 further comprising receiving elevation information from an elevation information source.
Apparatus for generating a three-dimensional site model of a site that contains objects comprising:
- (a) means for providing at least one of aerial imagery and oblique imagery of the site;
- (b) means for processing at least one of the aerial imagery or the oblique imagery to generate a plan view model comprising object outlines,
- wherein processing fun her comprises means for producing an elevation map using key video images;
- means for analyzing the elevation map to identify surface patches;
- means for computing local orientation of the surface patches within the elevation map;
- means for aggregating the surface patches for the key video images with respect to the local orientation of the surface patches;
- means for aggregating elevation information amongst the key video images to produce a three-dimensional structure;
- and means for processing the three-dimensional structure into a plan view model;
- and
- (c) means for producing from the plan view a three-dimensional site model comprising blocks that represent objects within the site,
- wherein the objects within the site include at least one building the rectilinear outline of the at least one building being determined by:
- computing an orientation histogram for the initial rectangle representing the building and selecting the angle of orientation with die strongest peak as the dominant orientation of the building;
- computing a gradient field based on image intensity of the building;
- integrating the gradient field along and perpendicular to the dominant orientation of the building;
- fitting straight lines to local maximums of the gradient field along each direction;
- detecting knot points at the crossings of the straight lines;
- dividing the initial rectangle into a plurality of smaller rectangles based on the knot points;
- and
- traversing the smaller rectangles to find the rectilinear outline of the building.
The apparatus of clam 17 wherein means (d) further comprises:
- (d1) means for selecting a building within the site;
- (d2) means for detecting edges of the building within the oblique imagery;
- (d3) means for identifying a set of roof model hypotheses that correspond to the edges;
- (d4) means for selecting and fitting each roof model hypothesis to the oblique imagery to identify a roof model hypothesis as the roof model.
Dependencies are inferred from the claim text.