Method and apparatus for producing digital orthophotos using sparse stereo configurations and external models

• US 6,757,445 B1
• Filed: 10/04/2000
• Issued: 06/29/2004
• Est. Priority Date: 10/04/2000
• Status: Expired due to Term

• Current Assignee: Quantum Spatial Inc.
• Original Assignee: Pixxures Inc. (Xedar Corporation)
• Inventors: Knopp, David E.
• Primary Examiner(s): Johnson, Timothy M.
• Assistant Examiner(s): Bayat, Ali
• Application Number:
  US09/679,726
• Time in Patent Office: 1,364 Days
• Field of Search: 382/106, 382/154, 382/285, 382/296, 382/300, 382/284, 382/294, 382/289, 345/421, 345/582, 345/649, 345/653, 345/620, 345/629, 345/630, 345/634, 345/619, 345/672, 345/679, 345/677, 348/135, 348/144, 348/47, 396/7, 396/8, 396/9, 396/10, 396/11, 396/13, 356/2, 358/537, 358/540
• US Class Current: 382/285
• CPC Class Codes: G01C 11/06 by comparison of two or more pictures of the same area

Abstract

A method and system for producing digital orthophotos from imagery acquired as full or sparse stereo. The orthophotos can be produced in a variety of map coordinate systems without the need to convert or recompute DEM or photogrammetric solution data. In one embodiment, a two dimensional, planimetric free-network solution, utilizing arbitrary datum definition constraints, is used to provide a transitory coordinate system that is used to facilitate the image measurement process. It is utilized as a preliminary step to refine apriori block layout information to facilitate point picking and to provide general quality control capabilities before undertaking a rigorous 3D photogrammetric adjustment. In place of a general map conversion transformation, an identity transformation can be used, so that map coordinates and world coordinates are identical. With this process, given DEM data and photogrammetric solution data in a particular coordinate system, the orthophoto image data can be produced in any map coordinate system. In one embodiment, all geometric coordinate transformations are performed prior to performing the image intensity interpolation operation. Thus, only one image intensity interpolation operation is performed, using the geometric coordinate data. In another embodiment, a network constraint is introduced to the block adjustment process that assumes an average vertical direction in order to support the process of self rectification.

Claims

1. A method for improving the quality of tie points and control points measured in producing a digital orthophoto from a set of overlapping images of a project area, said method comprising the steps of:

measuring a plurality of points within a block of images of the project area;

performing a two-dimension, free network adjustment using the points measured within the block of images to simultaneously compute a two-dimension orientation transformation for each image in the block of images;

performing a three-dimension, free-network adjustment using the results of the two-dimension, free-network adjustment;

assessing the quality of the points for the block of images using the results of the three-dimension, free-network adjustment;

refining the measurement of at least one of the points; and

performing at least a further three-dimension, free-network adjustment.

2. The method according to claim 1, wherein the step of measuring points includes displaying overlapping images in pairs, measuring only corner points in each image pair, subsequently measuring additional points, and performing said two-dimension, free-network adjustment prior to measuring said additional points.

3. The method according to claim 1, wherein assessing the quality of the points for the block of images includes assessing residuals of the points after performing said three-dimension, free-network adjustment.

4. A method for measuring tie points and control points in producing a digital orthophoto from a block of overlapping images of a project area, said method comprising the steps of:

displaying the overlapping images in pairs,

measuring corner points in each image pair to obtain point data;

performing a two-dimension, free network adjustment using the point data obtained by measuring points within the block of images to simultaneously compute a two-dimension orientation transformation for each image in the block; and

measuring a plurality of additional tie points within the block of images of the project area, wherein said two-dimension, free-network adjustment is performed prior to measuring said additional points.

5. A method for providing elevation model data for use in producing a digital orthophoto of a project area;

said method comprising the steps of;

acquiring elevation model data from at least first and second sources;

prioritizing the elevation model data acquired from the first and second sources;

storing the elevation model data for at least the first and second elevation models in a memory in a predetermined order to produce an elevation multimodel;

accessing the elevation model data having the highest priority to provide elevation data for use in performing a first geometric transformation; and

accessing the elevation model data having the next priority to provide elevation data for use in performing a second geometric transformation.

6. The method according to claim 5, wherein the elevation model data for the first and second elevation models are stored in contiguous data storage locations within the memory.

7. The method according to claim 5, wherein the elevation multimodel is used as a source of external control to constrain a photogrammetric adjustment.

8. The method according to claim 5, wherein the elevation multimodel is used as a source of external control in an orthorectification process.

9. The method according to claim 5, wherein the elevation multimodel is used as a source of one-dimension elevation model data in deriving full three-dimensional data from a two dimension map and a one-dimension elevation model.

10. The method according to claim 5, wherein the elevation multimodel is used as a source of external control to constrain a photogrammetric adjustment process.

11. A method for producing a digital orthophoto from a block of overlapping images of a project area;

said method comprising the steps of;

acquiring imagery data for the project area;

performing a plurality of geometric coordinate transformations on the imagery data to obtain geometric coordinate data, wherein the step of performing a plurality of geometric coordinate transformations includes performing at least one photogrammetric orientation and performing at least one cartographic transformation; and

performing only one image intensity interpolation operation using the geometric coordinate data, wherein all of said geometric coordinate transformations are performed prior to performing said image intensity interpolation operation.

12. A method for producing a digital orthophoto from a block of overlapping images of a project area;

said method comprising the steps of;

obtaining imagery data expressed in cartographic coordinates;

transforming the cartographic coordinates to world coordinates to obtain imagery data expressed in world coordinate data;

using the world coordinate data to obtain elevation value data from an elevation model expressed in world coordinates;

computing a photogrammetric orientation solution using the elevation value data obtained from the elevation model and the imagery data expressed in world coordinates;

performing a pixel intensity interpolation operation using the results of the photogrammetric solution and perspective image data; and

using the results of the pixel intensity interpolation operation and the imagery data expressed in cartographic coordinates to produce an orthophoto expressed in cartographic coordinates.

13. The method according to claim 12, wherein the elevation model includes a first source of elevation model data and a second source of elevation model data;

and including the steps of prioritizing the accessing of the first and second sources of elevation model data; and

accessing the first and second sources of elevation model data in the order of priority.

14. A method for producing digital orthophoto from a block of overlapping images of a project area, said method comprising the steps of:

acquiring digital imagery data for the project area by taking a series of aerial photographs in sparse stereo configuration wherein the imagery data comprises a plurality of frames wherein adjacent frames overlap by no more than about twenty percent in each/at least one direction;

selecting a plurality of points in overlapping portions of adjacent frames to obtain control data;

computing a photogrammetric solution using the imagery data and the control data; and

performing a rectification operation using the results of the photogrammetric solution.

15. The method according to claim 14, wherein at least performing the rectification operation includes obtaining elevation data from a sparse stereo elevation model.

16. A method for producing a digital orthophoto for a project area, said method comprising the steps of:

using an uncalibrated camera to obtain a series of aerial photographs of the project area for providing digital imagery representing a block of overlapping images of the project area;

introducing at least one perspective camera model parameter into a bundle block adjustment program as an unknown parameter pertaining to the taking camera;

performing a bundle block adjustment operation using the digital imagery data to calculate said unknown parameter;

supplying the calibrated parameter data and bundle adjustment result data to an orthorectification process; and

performing an orthorectification operation using the calibrated parameter data and the bundle adjustment data.

17. The method according to claim 16, wherein the aerial photo configuration includes those with sparse stereo.

18. The method according to claim 16, wherein the perspective camera model parameters include exterior orientation parameters and interior orientation parameters.

19. The method according to claim 16, wherein the exterior orientation parameters include principal point-x, principal point-y and principal distance.

20. A method for creating orthophoto images from imagery without stereo overlap and without availability of an external elevation model, comprising:

performing a photogrammetric adjustment using tie points and control points for the sparse stereo geometry to provide a three dimensional coordinate value for a plurality of ground points which correspond to all points which have been measured in two or more images, and wherein the ground points form a sparse sampling of the ground elevation model;

each ground point having a known elevation value expressed in world space and a planimetric location expressed in world space; and

interpolating an elevation model between the ground points for use in producing a digital orthophoto for a project area.

21. The method according to claim 20, including using a Delaunay triangulation for interpolating the said elevation model between the ground points.

22. The method according to claim 20, including said elevation model between the ground points.

23. A method for producing a digital orthophoto from a block of overlapping images for a project area, said method comprising the steps of:

generating a map footprint which represents the outer edge of meaningful pixel intensity values expressed in a coordinate system related to internal geometry of the taking camera;

generating a three-dimension world sample grid using an elevation model, the three-dimension world sample grid including a grid of points representing a point sampling of the elevation model consistent with a regular and homogenous spacing in a map coordinate system; and

generating a pixel displacement grid wherein every pixel location of the output image has a look-up value that indicates from which source image location to extract radiometric information.

24. The method according to claim 23, wherein the step of generating a map footprint includes the steps of converting the data boundary from pixel coordinates to camera coordinates, back projecting the data boundary into world space, intersecting boundary directions with the elevation model to produce a world footprint, and converting the world footprint to a map footprint.

25. The method according to claim 23, wherein the step of generating a three-dimension world sample grid, includes the steps of determining maximum and minimum map coordinate extents, generating a sample grid in map coordinates, converting the map grid to a world grid, and assigning elevation values at world grid points using the elevation model.

26. The method according to claim 23, wherein the step of generating a pixel displacement grid, includes the steps of forward projecting the world grid into image coordinates, converting the image coordinates to pixel coordinates, computing displacements at each location, and interpolating the displacement grid at each pixel location.

27. A method for producing a digital orthophoto from a block of overlapping images for a project area, said method comprising the steps of:

generating a map footprint which represents the outer edge of meaningful pixel intensity values expressed in a coordinate system related to internal geometry of the taking camera, wherein generating the map footprint includes converting the data boundary from pixel coordinates to camera coordinates;

back projecting the data boundary into world space;

intersecting boundary directions with an elevation model to produce a world footprint; and

converting the world footprint to a map footprint;

generating a three-dimension world sample grid, including determining maximum and minimum map coordinate extents;

generating a sample grid in map coordinates;

converting the map grid to a world grid; and

assigning elevation values at world grid points using the elevation model; and

generating a pixel displacement grid, including forward projecting the world grid into image coordinates;

converting the image coordinates to pixel coordinates;

computing displacements at each location; and

interpolating the displacement grid at each pixel location.

28. A method for producing a digital orthophoto from a block of overlapping images for a project area, said method comprising the steps of:

acquiring digital imagery data representing the block of overlapping images of the project area;

obtaining measurement data for tie and control points;

assuming vertical direction for the imagery;

applying an assumed vertical direction constraint to a block bundle adjustment program;

performing a block adjustment using the block bundle adjustment program;

deriving elevation model data from the imagery data;

performing an orthorectification operation using the derived elevation model data; and

transforming the geoimage to a map coordinate system.

29. A method for producing a digital orthophoto from a block of overlapping images for a project area, said method comprising the steps of:

acquiring digital imagery data representing the block of overlapping images of the project area;

obtaining measurement data for a plurality of tie and a plurality of control points;

performing photogrammetric interpolation adjustment using a block bundle adjustment process to obtain an elevation interpolation model including three-dimension ground points;

fitting the elevation interpolation model between ground points; and

interpolating elevation at arbitrary ones of the three-dimension ground points to support rectification in world coordinates.

30. The method according to claim 29 including applying an assumed vertical direction constraint to the block bundle adjustment process prior to performing the photogrammetric interpolation adjustment.