Automated analytic stereo comparator

• US 5,606,627 A
• Filed: 01/24/1995
• Issued: 02/25/1997
• Est. Priority Date: 01/24/1995
• Status: Expired due to Fees

• Current Assignee: Eotek, Inc.
• Original Assignee: Eotek, Inc.
• Inventors: Kuo, Shihjong
• Primary Examiner(s): Boudreau, Leo
• Assistant Examiner(s): KELLEY, CHRISTOPHER S
• Application Number:
  US08/378,257
• Time in Patent Office: 763 Days
• Field of Search: 382/106, 382/108, 382/109, 382/154, 382/266, 382/294, 348/42, 348/144, 348/159, 356/2, 356/3, 356/12, 33/20.4, 354/65, 354/68, 354/113
• US Class Current: 382/154
• CPC Class Codes: G01C 11/06 by comparison of two or more pictures of the same area
• CPC Class Codes: G06T 2207/10012 Stereo images
• CPC Class Codes: G06T 2207/10032 Satellite or aerial image; Remote sensing
• CPC Class Codes: G06T 2207/30181 Earth observation
• CPC Class Codes: G06T 7/593 from stereo images
• CPC Class Codes: H04N 13/122 Improving the 3D impression of stereoscopic images by modifying image signal contents, e.g. by filtering or adding monoscopic depth cues H04N13/128 takes precedence
• CPC Class Codes: H04N 13/189 Recording image signals; Reproducing recorded image signals
• CPC Class Codes: H04N 13/221 using the relative movement between cameras and objects
• CPC Class Codes: H04N 2013/0081 Depth or disparity estimation from stereoscopic image signals
• CPC Class Codes: H04N 2013/0096 Synchronisation or controlling aspects

Abstract

A method and an apparatus are disclosed that extract digital elevation data from a pair of stereo images with two corresponding sets of airborne control data associated with each image of the stereo image pair. Point and edge features are identified from the stereo image pair, projected from respective camera stations of the stereo images onto a horizontal projection plane using respective set of airborne control dam including respective positions of a camera station, the interior orientation and calibration data of the camera, respective roll angles, respective pitch angles, respective yaw angles of each image, and respective flight bearing angles. The space coordinate of each projection are compared in object space to obtain an object-space parallax. The topographic elevation of each feature is derived from said object-space parallax, a base length, and the altitude of a camera station autonomously to eliminate the need for stereoscopic viewing device or ground control data.

Claims

1. A method of extracting topographic elevation from two partially overlapping images of terrain and two corresponding sets of airborne control data including respective camera stations, a set of camera parameters, respective roll angles, respective pitch angles, respective yaw angles, respective flight bearing angle, associated with each image stored in a random access memory device comprising the steps of:

(a) identifying point and edge features in both the left image and the right image of said pair of images,

(b) constructing a list of common features containing a first list of left photograph coordinates of common features in said left image and a second list of right photograph coordinates of common features in said right image,

(c) projecting a common feature point in said left image from a left camera station onto a horizontal plane at a height h above a reference Datum plane,

(d) calculating a first space coordinate of the projection of said common feature point onto said horizontal plane from said left image using the first set of airborne control data associated with said left image,

(e) projecting said common feature point in said right image from a right camera station onto said horizontal plane,

(f) calculating a second space coordinate of the projection of said common feature point onto said horizontal plane from said right image using the second set of airborne control data associated with said right image,

(g) calculating an object space parallax between said first space coordinate and said second space coordinate,

(h) estimating the true elevation h'"'"' of said common feature point from said object-space parallax, a base length, and the attitude of said left camera station,

(i) repeating the steps (c) and (d) where said feature point is projected onto a new horizontal plane at height h'"'"' above said reference Datum to obtain a best three dimensional space coordinate of said feature point at elevation h'"'"',

(j) repeating the steps (c) through (i) for each common feature point in said list of common features.

2. The method of claim 1, wherein there is further included the step of:

adjusting both sets of roll angles, pitch angles, yaw angles, flight bearing angles for said left image and said right image according to a set of solutions of small angle correction terms to said first set of roll angle, pitch angle, yaw angle, bearing angle from step (d) and to said second set of roll angle, pitch angle, yaw angle, bearing angle from step (f).

3. The method of claim 1, wherein in step (a) the step of identifying includes:

filtering both said left image and said right image with edge enhancing band pass filters.

4. The method of claim 1, wherein in step (b) the step of constructing includes:

evaluating image correlation on respective image patches from said left image and said right image to identify point and edge features common to both images.

5. The method of claim 1, wherein there is further included the step of:

storing a list of said best three dimensional space coordinates referenced to said reference Datum plane for all common feature points in said list of common features from step (j).

6. A method of extracting topographic elevation from two partially overlapping images of terrain and two corresponding sets of airborne control data including respective camera station, a set of camera parameters, respective roll angle, respective pitch angle, respective yaw angle, respective flight bearing angle, associated with each image stored in an unprocessed mission media comprising the steps of:

(a) retrieving both images and said both sets of airborne control data from said mission medium into a random access memory device,

(b) identifying point and edge features in both the left image and the right image of said pair of images,

(c) constructing a list of common features containing a first list of left photograph coordinates of common features in said left image and a second list of right photograph coordinates of common features in said right image,

(d) projecting a common feature point in said left image from a left camera station onto a horizontal plane at a height h above a reference Datum plane,

(e) calculating a first space coordinate of the projection of said common feature point onto said horizontal plane from said left image using the first set of airborne control data associated with said left image,

(f) projecting said common feature point in said right image from a right camera station onto said horizontal plane,

(g) calculating a second space coordinate of the projection of said common feature point onto said horizontal plane from said right image using the second set of airborne control data associated with said right image,

(h) calculating an object space parallax between said first space coordinate and said second space coordinate,

(i) estimating the true elevation h'"'"' of said common feature point from said object-space parallax, a base length, and the attitude of said left camera station,

(j) repeating the steps (d) and (e) where said feature point is projected onto a new horizontal plane at height h'"'"' above said reference Datum to obtain a best three dimensional space coordinate of said feature point at elevation h'"'"',

(k) repeating the steps (d) through (j) for each common feature point in said list of common features.

7. The method of claim 6, wherein there is further included the step of:

adjusting both sets of roll angles, pitch angles, yaw angles, bearing angles for said left image and said right image according to a set of solutions of small angle correction terms to said first set of roll angle, pitch angle, yaw angle, bearing angle from step (e) and to said second set of roll angle, pitch angle, yaw angle, bearing angle from step (g).

8. The method of claim 6, wherein in step (b) the step of identifying includes:

filtering both said left image and said right image with edge enhancing band pass filters.

9. The method of claim 6, wherein in step (c) the step of constructing includes:

evaluating image correlation on respective image patches from said left image and said right image to identify point and edge features common to both images.

10. The method of claim 6, wherein there is further included the step of:

storing a list of said best three dimensional space coordinates referenced to said reference Datum plane for all common feature points in said list of common features from step (k).

11. A stereocomparator for extracting topographical elevation information from digital image data and related airborne control data stored in a unprocessed mission media comprising:

(a) a storage device containing said mission data medium storing at least one set of two partially overlapping images of terrain and two corresponding sets of airborne control data including respective camera station, a set of camera parameters, respective roll angle, respective pitch angle, respective yaw angle, respective flight bearing angle, associated with each image,

(b) an input means for retrieving said pair of images and said two corresponding sets of airborne control data into a random access memory device,

(c) a feature extraction means for identifying point and edge features in both the left image and the right image of said pair of images,

(d) a common list construction means for constructing a list of common features containing a first list of left photograph coordinates of said common features in said left image and a second list of right photograph coordinates of said common features in said right image,

(e) a projection means for projecting a common feature point in said left image from a left camera station onto a horizontal plane at a height h above a reference Datum plane and calculating a first space coordinate from left image projection, and projecting said common feature point in said right image from a right camera station onto said horizontal plane and calculating a second space coordinate from right image projection,

(f) a coordinate comparator means for calculating an object space parallax between said first space coordinate and said second space coordinate,

(g) a elevation estimator means for estimating the true elevation of said common feature point from said object space parallax, a base length, and the attitude of said left camera station.

12. A stereocomparator as defined in claim 11, wherein said projection means further includes an attitude adjustment means for:

adjusting both sets of roll angles, pitch angles, yaw angles, bearing angles for said left image and said right image according to a set of solutions of small angle correction terms to said sets of roll angles, pitch angles, yaw angles, bearing angles.

13. A stereocomparator as defined in claim 11, wherein said feature extraction means includes a band-pass filtering means for:

filtering both said left image and said right image to extract point and edge features.

14. A stereocomparator as defined in claim 11, wherein further includes an output means for:

storing a list of three dimensional space coordinates referenced to said reference Datum plane.