Automatic feature point detection for calibration of multi-camera systems

• US 10,643,091 B2
• Filed: 08/16/2019
• Issued: 05/05/2020
• Est. Priority Date: 10/30/2015
• Status: Active Grant

• Current Assignee: Texas Instruments, Inc.
• Original Assignee: Texas Instruments, Inc.
• Inventors: Kwon, Do-Kyoung, Batur, Aziz Umit, Appia, Vikram
• Primary Examiner(s): Saini, Amandeep
• Application Number:
  US16/542,368
• Publication Number:
  US 20190370589A1
• Time in Patent Office: 263 Days
• Field of Search: 382199
• CPC Class Codes: G06T 2207/30204 Marker
• CPC Class Codes: G06T 2207/30248 Vehicle exterior or interior
• CPC Class Codes: G06T 7/80 Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration
• CPC Class Codes: G06V 10/245 by locating a pattern; Special marks for positioning
• CPC Class Codes: G06V 10/462 Salient features, e.g. scale invariant feature transforms [SIFT]
• CPC Class Codes: H04N 23/698 for achieving an enlarged field of view, e.g. panoramic image capture

Abstract

In some embodiments, a computer-readable medium stores executable code, which, when executed by a processor, causes the processor to: capture an image of a finder pattern using a camera; locate a predetermined point within the finder pattern; use the predetermined point to identify multiple boundary points on a perimeter associated with the finder pattern; identify fitted boundary lines based on the multiple boundary points; and locate feature points using the fitted boundary lines.

Claims

1. A method, comprising:

capturing an image using a first camera in a multi-camera system, the image containing a representation of a finder pattern;

locating a first point in the finder pattern;

locating boundary points along a perimeter of the finder pattern using the first point;

fitting boundary lines along the perimeter of the finder pattern using the boundary points;

locating feature points along the perimeter using the fitted boundary lines; and

generating a perspective transformation matrix based on the feature points.

2. The method of claim 1, wherein the first point of the finder pattern is at a center of the finder pattern.

3. The method of claim 1, wherein the finder pattern comprises a first square centered within a second square.

4. The method of claim 3, further comprising binarizing the image, wherein, after the binarization is complete, one of the first and second squares is white and the other of the first and second squares is black.

5. The method of claim 4, wherein the binarized image includes the finder pattern positioned on an object, the object has a color matching that of the first square.

6. The method of claim 1, further comprising using the perspective transformation matrix to generate a bird'"'"'s-eye view of an object on which the multi-camera system is mounted.

7. The method of claim 1, wherein the feature points are points at which the fitted boundary lines intersect.

8. The method of claim 1, wherein locating the first point of the finder pattern comprises identifying a point in the finder pattern that lies along a line that intersects each perimeter of the finder pattern twice, wherein the line has segments, each segment located between a different pair of consecutive perimeter crossings, and wherein a ratio of lengths of the segments falls within a predetermined ratio range.

9. The method of claim 1, wherein locating the boundary points comprises identifying points at which a line passing through the predetermined point intersects the perimeter of the finder pattern.