Intrinsic feature-based pose measurement for imaging motion compensation

• US 8,810,640 B2
• Filed: 05/16/2011
• Issued: 08/19/2014
• Est. Priority Date: 05/16/2011
• Status: Active Grant

• Current Assignee: UT-Battelle LLC
• Original Assignee: UT-Battelle LLC
• Inventors: Baba, Justin S., Goddard, James Samuel Jr.
• Primary Examiner(s): Patel, Jay
• Assistant Examiner(s): Aghevli, Reza
• Application Number:
  US13/108,395
• Publication Number:
  US 20120293667A1
• Time in Patent Office: 1,191 Days
• Field of Search: None
• US Class Current: 348/68
• CPC Class Codes: G06T 2207/10021 Stereoscopic video; Stereoscopic image sequence
• CPC Class Codes: G06T 2207/10108 Single photon emission computed tomography [SPECT]
• CPC Class Codes: G06T 2207/20076 Probabilistic image processing
• CPC Class Codes: G06T 2207/30004 Biomedical image processing
• CPC Class Codes: G06T 7/251 involving models
• CPC Class Codes: G06T 7/277 involving stochastic approaches, e.g. using Kalman filters
• CPC Class Codes: G06T 7/75 involving models
• CPC Class Codes: H04N 13/246 Calibration of cameras
• CPC Class Codes: H04N 2013/0085 Motion estimation from stereoscopic image signals
• CPC Class Codes: H04N 7/00 Television systems details H04N3/00, H04N5/00; methods or arrangements, for coding, decoding, compressing or decompressing digital video signals H04N19/00; selective content distribution H04N21/00
• CPC Class Codes: H04N 7/01 Conversion of standards , e.g. involving analogue television standards or digital television standards processed at pixel level

Abstract

Systems and methods for generating motion corrected tomographic images are provided. A method includes obtaining first images of a region of interest (ROI) to be imaged and associated with a first time, where the first images are associated with different positions and orientations with respect to the ROI. The method also includes defining an active region in the each of the first images and selecting intrinsic features in each of the first images based on the active region. Second, identifying a portion of the intrinsic features temporally and spatially matching intrinsic features in corresponding ones of second images of the ROI associated with a second time prior to the first time and computing three-dimensional (3D) coordinates for the portion of the intrinsic features. Finally, the method includes computing a relative pose for the first images based on the 3D coordinates.

Claims

1. A method for generating pose information for motion correcting images, comprising the steps of:

obtaining a first plurality of images of a region of interest (ROI) to be imaged acquired using a pair of cameras located at different positions and orientations with respect to the ROI, the first plurality of images associated with a first time and define one or more stereo image pairs of the ROI;

defining active regions in each of the first plurality of images,

selecting intrinsic features in each of the first plurality of images based on the corresponding ones of the active regions,

identifying a portion of the intrinsic features temporally and spatially matching intrinsic features in corresponding ones of a second plurality of images of the ROI associated with a second time prior to the first time,

calculating three-dimensional (3D) coordinates for the portion of the intrinsic features based on an optical triangulation using the stereo image pairs, the optical triangulation based at least on image coordinates for each of the pair of cameras, image disparity associated with the stereo image pairs, and a physical configuration of the pair of cameras; and

computing a relative pose of the ROI in 3D space for the first plurality of images based on the 3D coordinates for the portion of the intrinsic features.

2. The method of claim 1, wherein the step of obtaining comprises:

rectifying a plurality of raw images acquired using a plurality of cameras at the different positions to a canonical arrangement based on calibration parameters for the plurality of cameras to yield the first plurality of images.

3. The method of claim 2, wherein the step of obtaining further comprises:

prior to the rectifying, adjusting image attributes for the plurality of raw images.

4. The method of claim 2, wherein the rectifying comprises:

applying an image rectification transform to each of the plurality of raw images based on the calibration parameters to yield a plurality of rectified images that lie in a same plane, have a same focal length, and have a same orientation.

5. The method of claim 2, wherein the rectifying comprises:

applying an image rectification transform to the plurality of raw images based on the calibration parameters to yield a plurality of rectified images so that each of the intrinsic features in the plurality of images lies along at least one of a common horizontal line and a common vertical line.

6. The method of claim 1, wherein the defining the active region further comprises:

selecting as the active regions visible portions of the ROI extracted from the first plurality of images.

7. The method of claim 1, wherein the selecting the intrinsic features further comprises:

selecting as the intrinsic features a portion of discrete features extracted from the one of the first plurality of images located in the active region.

8. A system for generating pose information for motion correcting images, comprising:

a processor;

a non-transitory computer-readable storage medium storing a computer program for causing the processor to perform operations comprising;

obtaining a first plurality of images of a region of interest (ROI) to be imaged acquired using a pair of cameras located at different positions and orientations with respect to the ROI, the first plurality of images associated with a first time and define one or more stereo image pairs;

defining active regions in each of the first plurality of images,

selecting intrinsic features in each of the first plurality of images based on the corresponding ones of the active regions,

identifying a portion of the intrinsic features temporally and spatially matching intrinsic features in corresponding ones of a second plurality of images of the ROI associated with a second time prior to the first time,

calculating three-dimensional (3D) coordinates for the portion of the intrinsic features based on an optical triangulation using the stereo image pairs, the optical triangulation based at least on image coordinates for each of the pair of cameras, image disparity associated with the stereo image pairs, and a physical configuration of the pair of cameras, and

computing a relative pose of the ROI in 3D space for the first plurality of images based on the 3D coordinates for the portion of the intrinsic features.

9. The system of claim 8, wherein the obtaining further comprising rectifying a plurality of raw images acquired using a plurality of cameras at the different positions to a canonical arrangement based on calibration parameters for the plurality of cameras to yield the first plurality of images.

10. The system of claim 9, wherein the obtaining further comprises adjusting image attributes for the plurality of raw images prior to the rectifying.

11. The system of claim 9, wherein rectifying the plurality of raw images comprises:

applying an image rectification transform to each of the plurality of raw images based on the calibration parameters to yield a plurality of rectified images that lie in a same plane, have a same focal length, and have a same orientation.

12. The system of claim 9, wherein the rectifying the plurality of raw images comprises:

applying an image rectification transform to each of the plurality of raw images based on the calibration parameters to yield a plurality of rectified images.

13. The system of claim 8, wherein the active regions are defined by selecting as the active regions visible portions of the ROI extracted from the first plurality of images.

14. The system of claim 8, wherein the intrinsic features are selected by selecting as the intrinsic features a portion of discrete features extracted from the one of the first plurality of images located in the active region.

15. A non-transitory computer-readable medium storing a computer program having instructions for generating pose information for motion correcting images, the instructions comprising:

obtaining a first plurality of images of a region of interest (ROI) to be tomographically imaged and associated with a first time, the first plurality of images acquired via a pair of cameras located at with different positions and orientations with respect to the ROI and define one or more stereo image pairs;

processing each one of the first plurality of images, the processing comprising;

defining an active region in the one of the first plurality of images,

selecting intrinsic features in the one of the first plurality of images based on the active region,

identifying a portion of the intrinsic features temporally and spatially matching intrinsic features in a corresponding one of a second plurality of images of the ROI associated with a second time prior to the first time, and

computing three-dimensional (3D) coordinates for the portion of the intrinsic features based on an optical triangulation using the stereo image pairs, the optical triangulation based at least on image coordinates for each of the pair of cameras, image disparity associated with the stereo image pairs, and a physical configuration of the pair of cameras; and

computing a relative pose of the ROI in 3D space for the first plurality of images based on the 3D coordinates for the portion of the intrinsic features.

16. The non-transitory computer-readable medium of claim 15, wherein the obtaining comprises:

adjusting image attributes for a plurality of raw images of the ROI acquired using a plurality of cameras at the different positions; and

rectifying the plurality of raw images to a canonical arrangement based on calibration parameters for the plurality of cameras to yield the first plurality of images.

17. The non-transitory computer-readable medium of claim 16, wherein the rectifying comprises:

applying an image rectification transform to each of the plurality of raw images based on the calibration parameters to yield a plurality of rectified images that lie in a same plane, have a same focal length, and have a same orientation.

18. The non-transitory computer-readable medium of claim 16, wherein the rectifying comprises:

applying an image rectification transform to the plurality of raw images based on the calibration parameters to yield a plurality of rectified images so that each of the intrinsic features in the plurality of images lies along at least one of a common horizontal line and a common vertical line.

19. The non-transitory computer-readable medium of claim 15, wherein the defining the active region further comprises:

selecting as an outline of the active region a visible portion of the ROI extracted from the one of the first plurality images.

20. The non-transitory computer-readable medium of claim 15, wherein the selecting the intrinsic features further comprises:

selecting as the intrinsic features a portion of discrete features extracted from the one of the first plurality of images located in the active region.