Method and apparatus for determining points of interest on an image of a camera calibration object

• US 6,768,509 B1
• Filed: 06/12/2000
• Issued: 07/27/2004
• Est. Priority Date: 06/12/2000
• Status: Expired due to Term

• Current Assignee: Beijing Xiaomi Mobile Software Co., Ltd. (Xiaomi Corp.)
• Original Assignee: Intel Corporation
• Inventors: Bouguet, Jean-Yves, Bradski, Gary R., Mosyagin, Valery, Pisarevsky, Vadim
• Primary Examiner(s): MOE, AUNG SOE
• Application Number:
  US09/592,071
• Time in Patent Office: 1,506 Days
• Field of Search: 348/207.99, 348/222.1, 348/252, 348/253, 348/254, 348/175, 348/179, 348/180, 348/181, 348/182, 348/187, 348/188, 382/209, 382/194, 358/504
• US Class Current: 348/207.99
• CPC Class Codes: G06T 7/12 Edge-based segmentation
• CPC Class Codes: G06T 7/155 involving morphological operators
• CPC Class Codes: G06T 7/194 involving foreground-background segmentation
• CPC Class Codes: G06T 7/85 Stereo camera calibration
• CPC Class Codes: G06V 10/46 Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features colour feature extraction G06V10/56
• CPC Class Codes: H04M 17/02 Coin-freed or check-freed systems , e.g. mobile- or card-operated phones, public telephones or booths
• CPC Class Codes: H04N 13/246 Calibration of cameras

Abstract

The present invention allows for the locations of points of interest in a calibration object in a calibration image for a digital camera to be identified automatically. The image is an array of pixels corresponding to the calibration object, which has a known reference pattern. In a preferred embodiment, the invention includes receiving an array of pixels produced by a camera, classifying each pixel of the image as light or dark, extracting contours from the image by identifying lines between light and dark pixels, comparing the extracted contours to the shapes of the known reference pattern, and identifying the shapes of the known reference pattern in the image using the extracted contours. Preferably, the image is a color image, and the color information in the pixels of the image are converted into gray scale values to render the image as a gray scale image before the pixels are classified. Classifying each pixel includes applying a brightness threshold to each pixel of the image to classify those pixels having a brightness greater than the threshold as light and those below the threshold as dark.

Claims

1. A method comprising:

receiving an array of pixels produced by a camera, the pixels corresponding to an image of a calibration object, the calibration object having a known reference pattern;

classifying each pixel of the image as light or dark;

extracting contours from the image by identifying lines between light and dark pixels;

comparing the extracted contours to the shapes of the known reference pattern; and

identifying the shapes of the known reference pattern in the image using the extracted contours.

2. The method of claim 1 wherein the image is a color image, the method further comprising converting color information in the pixels of the image into gray scale values to render the image as a gray scale image before classifying the pixels.

3. The method of claim 1 wherein classifying each pixel comprises applying a brightness threshold to each pixel of the image to classify those pixels having a brightness greater than the threshold as light and those below the threshold as dark.

4. The method of claim 3 wherein the brightness threshold is determined by the mean brightness value of the pixels in the image reduced by a portion of the highest brightness value of any pixel.

5. The method of claim 1 further comprising dilating the image after classifying each pixel.

6. The method of claim 1 wherein extracting the contours comprises applying the Douglas-Peucker algorithm to approximate a simple smooth shape.

7. The method of claim 6 wherein applying the Douglas-Peucker algorithm comprises drawing a connected arc along a contour that is confined to within a line threshold of being a line.

8. The method of claim 7 wherein applying the Douglas-Peucker algorithm comprises adaptively modifying the line threshold to obtain the calibration object shapes.

9. The method of claim 1 wherein identifying the contours comprises rejecting contours that correspond to contours on the calibration object that are smaller than a threshold size.

10. The method of claim 1 wherein identifying the contours comprises rejecting contours that correspond to contours on the calibration object that are distorted in shape by more than a threshold amount.

11. The method of claim 1 wherein the calibration object shapes are ordered quadrilaterals and wherein comparing the contours further comprises drawing perimeters around the extracted contours, examining the perimeters to identify quadrilaterals, rejecting perimeters that do not correspond to quadrilaterals, and finding corners for each perimeter, and wherein identifying the shapes further comprises ordering the corners to correspond to the quadrilaterals of the calibration object, and rejecting all identified quadrilaterals that are not within the ordering.

12. The method of claim 1 wherein the calibration object shapes are ordered quadrilaterals and wherein identifying the shapes further comprises extracting quadrilateral corner points from the extracted contours, drawing a perimeter around the extracted quadrilateral corner points, and identifying corners in the perimeter, and wherein identifying the shapes further comprises, counting the corners on each edge of the quadrilateral perimeter.

13. A machine-readable medium having stored thereon data representing sequences of instructions which, when executed by a processor, cause the processor to perform operations comprising:

receiving an array of pixels produced by a camera, the pixels corresponding to an image of a calibration object, the calibration object having a known reference pattern;

classifying each pixel of the image as light or dark;

extracting contours from the image by identifying lines between light and dark pixels;

comparing the extracted contours to the shapes of the known reference pattern; and

identifying the shapes of the known reference pattern in the image using the extracted contours.

14. The medium of claim 13 wherein the instruction further comprise instruction for dilating the image after classifying each pixel.

15. The medium of claim 13 wherein the instruction for extracting the contours comprise applying the Douglas-Peucker algorithm to approximate a simple smooth shape.

16. The medium of claim 13 wherein the calibration object shapes are ordered quadrilaterals and wherein the instruction for comparing the contours further comprise drawing perimeters around the extracted contours, examining the perimeters to identify quadrilaterals, rejecting perimeters that do not correspond to quadrilaterals, and finding corners for each perimeter, and wherein the instructions for identifying the shapes further comprise ordering the corners to correspond to the quadrilaterals of the calibration object, and rejecting all identified quadrilaterals that are not within the ordering.

17. An apparatus for identifying a calibration object in an image comprising:

means for receiving an array of pixels produced by a camera, the pixels corresponding to an image of a calibration object, the calibration object having a known reference pattern;

means for classifying each pixel of the image as light or dark;

means for extracting contours from the image by identifying lines between light and dark pixels;

means for comparing the extracted contours to the shapes of the known reference pattern; and

means for identifying the shapes of the known reference pattern in the image using the extracted contours.

18. The apparatus of claim 17 wherein the means for extracting the contours comprises means for applying the Douglas-Peucker algorithm to approximate a simple smooth shape.

19. The apparatus of claim 17 wherein the calibration object shapes are ordered quadrilaterals and wherein the means for comparing the contours further comprises means for drawing perimeters around the extracted contours, means for examining the perimeters to identify quadrilaterals, means for rejecting perimeters that do not correspond to quadrilaterals, and means for finding corners for each perimeter, and wherein the means for identifying the shapes further comprises means for ordering the corners to correspond to the quadrilaterals of the calibration object, and means for rejecting all identified quadrilaterals that are not within the ordering.