Method for determining ground line
A method of determining a ground line from an input image. The method includes determining a plurality of ground line candidates from the image, determining a certain band a central-line of which is a boundary between a G region and other regions in a Ground Building Sky (GBS) map of the image, and determining the ground line of the image by selecting a ground line candidate, among the plurality of ground line candidates, having the greatest extent crossing the certain band.
1. A method of determining a ground line of an image, comprising:
determining a plurality of ground line candidates from the image;
determining a certain band having a central-line being a boundary between a ground (G) region and other regions in a Ground Building Sky (GBS) map of the image; and
determining the ground line of the image by selecting a ground line candidate, among the plurality of ground line candidates, having the greatest number of points lying within the certain band.
2. The method of claim 1 , further comprising:
generating the GBS map of the image from the image.
3. The method of claim 1 , further comprising:
detecting a plurality of horizontal straight lines belonging to a B region of the image,
wherein the determining of the plurality of ground line candidates uses the plurality of horizontal straight lines.
4. The method of claim 3 , wherein the detecting comprises:
extracting a plurality of straight lines from the image;
detecting a plurality of horizontal straight lines from among the plurality of straight lines; and
selecting the plurality of horizontal straight lines belonging to the B region from among the plurality of horizontal straight lines.
5. The method of claim 4 , wherein the selecting the plurality of horizontal straight lines belonging to the B region comprises:
comparing the B region in the GBS map of the image with the plurality of horizontal straight lines; and
selecting the plurality of horizontal straight lines belonging to the B region, based on the comparing.
6. The method of claim 3 , further comprising:
clustering the plurality of horizontal straight lines into a plurality of groups,
wherein the determining of the plurality of ground line candidates is based on the clustered plurality of horizontal straight lines.
7. The method of claim 6 , wherein the clustering comprises:
clustering the plurality of horizontal straight lines into middle-phase groups, based on x coordinates of the plurality of horizontal straight lines, by using a mean shift algorithm;
eliminating an outlier of each of the middle-phase groups;
calculating a main direction of each of the middle-phase groups; and
generating the plurality of groups by merging the middle-phase groups having similar main directions.
8. The method of claim 6 , further comprising:
determining a vanishing point of each of the plurality of groups.
9. The method of claim 8 , further comprising:
determining a vertical boundary line between the plurality of groups.
10. The method of claim 9 , wherein the determining of the plurality of ground line candidates further comprises:
detecting a straight line that passes through a plurality of points of the vertical boundary line and the vanishing point of each of the plurality of groups.
11. The method of claim 1 , wherein the determining the certain band determines the certain band that has a predetermined width in a y coordinate direction, from a pixel of the boundary between the G region and the other regions in the GBS map.
12. The method of claim 1 , wherein the determining of the ground line of the image comprises:
calculating a number of pixels of each of the plurality of ground line candidates that lie within the certain band; and
determining the ground line candidate that has the greatest number of pixels lying within the certain band as the ground line of the image.
13. A computer readable recording device storing a program for implementing a method of any one of claims 1 - 12 .