LINE SEGMENT AND ARC DETECTION APPARATUS

20160335517

15/220167

July 26, 2016

An apparatus and method to detect a line segment or arc using Hough transform. A Hough transform unit performs contour extraction on brightness image data to generate contour image data, with pixels having a pixel value of 0 to 255, performs the Hough transform on points in the contour image data, and counts additional values represented by pixel values of points in the contour image data in a Hough table. The Hough transform unit performs contour extraction on first to third component data to generate first to third contour data with pixels having a pixel value of 0 to 255, performs the Hough transform on points in the first to third contour data, and counts additional values represented by pixel values of points in the first to third contour data in the Hough table. The detection unit comprehensively evaluates the counts to detect a line segment or arc.

MegaChips Corporation (Osaka, JP)

1. A detection apparatus for detecting an arc in a specified image, comprising: a feature quantity calculating part configured to determine a feature quantity of each point in the specified image; a Hough transform part configured to convert an arc that passes through an arbitrary point in the specified image to a value set of coordinates (a,b,r) in a Hough space by defining orthogonal coordinates with respect to a reference point in the specified image and representing an arbitrary arc in the specified image using a value set of coordinates (a, b) of its center in an orthogonal coordinate system and a radius r; a counting part configured to add a value weighted by a feature quantity of a point in the specified image in a storage entry corresponding to a value set of coordinates (a,b,r) in the Hough space to which the point in the specified image is converted by the Hough transform part; and a detection part configured to detect the arc in the specified image by identifying a value set of coordinates (a,b,r) for which the count exceeds a first threshold.

2. A detection method for detecting an arc in a specified image, comprising: a feature quantity calculating step to determine a feature quantity of each point in the specified image; a Hough transform step to convert an arc that passes through an arbitrary point in the specified image to a value set of coordinates (a,b,r) in a Hough space by defining orthogonal coordinates with respect to a reference point in the specified image and representing an arbitrary arc in the specified image using a value set of coordinates (a, b) of its center in an orthogonal eooKlinaie system and a radius r; a counting step to add a value weighted by a feature quantity of a point in the specified image in a storage entry corresponding to a value set of coordinates (a,b,r) in the Hough space to which the point in the specified image is converted by the Hough transform part; and a detection step to detect the arc in the specified image by identifying a value set of coordinates (a,b,r) for which the count exceeds a first threshold.

3. A detection apparatus for detecting an arc in a specified image in a color space, comprising: a generating part configured to generate from the specified image a plurality of component image images in the color space; a feature quantity calculating part configured to determine a feature quantity of each point in each component image; a Hough transform part configured to convert an arc that passes through an arbitrary point in each component image to a value set of coordinates (a,b,r) in a Hough space by defining orthogonal coordinates with respect to a reference point in each component image and representing an arbitrary arc in each component image using a value set of coordinates (a, b) of its center in an orthogonal coordinate system and a radius r; a counting part configured to sum a value weighted by a feature quantity of a point in each component image in a storage entry corresponding to a value set of coordinates (a,b,r) in the Hough space to which the point in each component image is converted by the Hough transform part; a comprehensive counting part configured to comprehensively evaluate the counts for all component images to determine a comprehensive count; and a detection part configured to detect the arc in the specified image by identifying a value set of coordinates (a,b,r) for which the comprehensive count exceeds a first threshold.

4. A detection method for detecting an arc in a specified image in a color space, comprising: a generating step to generate from the specified image a plurality of component images in the color space; a feature quantity calculating step to determine a feature quantity of each point in each component image; a Hough transform step to convert an arc that passes through an arbitrary point in each component image to a value set of coordinates (a,b,r) in a Hough space by defining orthogonal coordinates with respect to a reference point in each component image and representing an arbitrary arc in each component image using a value set of coordinates (a, b) of its center in an orthogonal coordinate system and a radius r; a counting step to add a value weighted by a feature quantity of a point in each component image in a storage entry corresponding to a value set of coordinates (a,b,r) in the Hough space to which the point in each component image is converted by the Hough transform part; a comprehensive counting step to comprehensively evaluate the counts for all component images to determine a comprehensive count; and a detection step to detect the arc in the specified image by identifying a value set of coordinates (a,b,r) for which the comprehensive count exceeds a first threshold.

5. The detection apparatus according to claim 1, wherein the feature quantity calculating part determines a contour intensity represented by the feature quantity of the point in the specified image by performing contour extraction.

6. The detection apparatus according to claim 5, wherein the counting part counts an additional value represented by the contour intensity of the point in the specified image after the contour extraction.

7. The detection apparatus according to claim 5, wherein the counting part quantizes and then counts the contour intensity of the point in the specified image after the contour extraction.

8. The line segment detection apparatus according to claim 1, further comprising a brightness image generator configured, if the specified image is a color image, to generate a brightness image based on the color image prior to a Hough transform being performed.

9. The detection apparatus according to claim 3, wherein the feature quantity calculating part determines a contour intensity represented by the feature quantity of the point in each component image by performing contour extraction.

10. The detection apparatus according to claim 9, wherein the counting part counts an additional value represented by the contour intensity of the point in each component image after the contour extraction.

11. The detection apparatus according to claim 9, wherein the counting part quantizes and then counts the contour intensity of the point in each component image after the contour extraction.

12. The detection apparatus according to claim 11, wherein the counting part quantizes contour intensities for different component images using different quantizing functions.

13. The detection apparatus according to claim 3, wherein the comprehensive counting part determines the comprehensive count by weighting and summing the counts of the component images.

06; 06

edspap.20160335517