CHARACTERIZING POINT CHECKING REGION SETTING APPARATUS AND METHOD, AND IMAGE STABILIZING APPARATUS INCLUDING THE SAME

• US 20130176449A1
• Filed: 12/13/2012
• Published: 07/11/2013
• Est. Priority Date: 01/11/2012
• Status: Active Grant

• Current Assignee: Hanwha Vision Company Limited (Hanwha Corporation)
• Original Assignee: Samsung Techwin Corporation (Hanwha Corporation)
• Inventors: BAE, Ju-Han, PARK, Sang-Ji, CHON, Je-Youl
• Granted Patent: US 8,988,534 B2
• Time in Patent Office: Days
• US Class Current: 348/208.99
• CPC Class Codes: G06T 2207/20081 Training; Learning
• CPC Class Codes: G06T 7/246 using feature-based methods, e.g. the tracking of corners or segments
• CPC Class Codes: G06V 10/462 Salient features, e.g. scale invariant feature transforms [SIFT]
• CPC Class Codes: H04N 23/00 Cameras or camera modules comprising electronic image sensors; Control thereof
• CPC Class Codes: H04N 23/6811 based on the image signal

Abstract

Provided is an image stabilizing apparatus and method for correcting an image that is shaken due to a movement of a camera. The image stabilizing apparatus includes a characterizing point checking region setting unit including: a sample frame extract unit which extracts a plurality of image frames obtained for a certain period of time in image data obtained by photographing an object; and a frame analyzing unit which detects a plurality of characterizing points in the extracted plurality of image frames, and sets a characterizing point checking region which is used to check characterizing points in a currently input image frame.

Claims

1. A characterizing point checking region setting unit comprising:

a sample frame extract unit which extracts a plurality of image frames, obtained for a certain period of time, from image data obtained by photographing an object; and

a frame analyzing unit which detects a plurality of characterizing points in the extracted plurality of image frames, and sets a characterizing point checking region, which is used to check characterizing points in a currently input image frame.

2. The characterizing point checking region setting unit of claim 1, wherein the plurality of image frames are obtained by continuously photographing a same object.

3. The characterizing point checking region setting unit of claim 1, wherein the image data input in the sample frame extract unit is obtained by a camera fixed on a position.

4. The characterizing point checking region setting unit of claim 1, wherein the frame analyzing unit comprises:

a characterizing point detector which receives the plurality of image frames from the sample frame extract unit, and detects the plurality of characterizing points in the plurality of image frames;

a characterizing point classification unit which classifies the plurality of characterizing points into a plurality of clusters for each of the image frames;

a center point detector which detects a centroid point of the characterizing points in the plurality of image frames; and

a check region determination unit which sets the characterizing point checking region including a major cluster among the plurality of clusters based on the centroid point.

5. The characterizing point checking region setting unit of claim 4, wherein a cluster of the plurality of clusters is classified as the major cluster if the cluster comprises 50% or more of the characterizing points, and as a minor cluster if the cluster comprises less than 50% of the characterizing points.

6. The characterizing point checking region setting unit of claim 5, wherein the center point detector calculates an average of a centroid point in a major cluster of each of the plurality of image frames and of a centroid point in a minor cluster of the each of the plurality of image frames, sums the average values, and divides the summed average values by a number of the plurality of image frames to detect the centroid point of the characterizing points in the plurality of image frames.

7. The characterizing point checking region setting unit of claim 4, wherein the check region determination unit defines the characterizing point checking region as a square or a circle.

8. The characterizing point checking region setting unit of claim 4, further comprising a checking region adjusting unit which determines whether the characterizing point checking region includes a standard level of characterizing points of the image frames extracted for the certain period of time or greater.

9. The characterizing point checking region setting unit of claim 8, wherein the standard level is 80% of the characterizing points.

15. An image stabilizing apparatus comprising:

the characterizing point checking region setting unit of claim 1; and

an image adjusting unit which sets the characterizing point checking region in an image frame that is currently input, compares the currently input image frame with a reference image that is preset, and adjusts the currently input image frame as much as a shaking amount when it is determined that the current image frame is shaken.

16. The image stabilizing apparatus of claim 15, further comprising reference image setting unit which extracts an image frame that is the least shaken among the plurality of image frames taken for a certain period of time, and sets the extracted image frame as the reference image

17. The image stabilizing apparatus of claim 16, wherein the reference image setting unit extracts the least shaken image frame by determining distances between a plurality of characterizing points and a center point in each of the image frames.

18. The image stabilizing apparatus of claim 17, wherein the reference image setting unit comprises:

a center point detector which receives the plurality of image frames and detects the center point in each of the image frames;

a characterizing point detector which detects the plurality of characterizing points in each of the image frames;

a frame average calculator which calculates a plurality of frame averages each of which is an average of the distances between the center point and the plurality of characterizing points in each of the image frames;

a frame comparison value calculator which calculates a plurality of frame comparison values each of which is obtained by summing up absolute values each of which is obtained by subtracting, from the frame average of each of the frame, a frame average of another image frame; and

a reference frame selector which selects an image frame having the smallest frame average value as the reference image.

10. A characterizing point checking region setting method comprising:

receiving a plurality of image frames captured for a certain period of time; and

detecting a plurality of characterizing points in the plurality of image frames, and setting a characterizing point checking region for checking characterizing points in a currently input image frame.

11. The characterizing point checking region setting method of claim 10, wherein the setting the characterizing point checking region comprises:

detecting a plurality of characterizing points in each of the plurality of image frames;

classifying the plurality of characterizing points into a plurality of clusters for each of the image frames;

detecting a representative centroid point representing the plurality of image frames; and

setting the characterizing point checking region including a major cluster among the plurality of clusters, based on the representative centroid point.

12. The characterizing point checking region setting method of claim 11, wherein a cluster of the plurality of clusters is classified as the major cluster if the cluster comprises 50% or more of the characterizing points, and as a minor cluster if the cluster comprises less than 50% of the characterizing points.

13. The characterizing point checking region setting method of claim 12, wherein the representative centroid point is detected by detecting centroid points of the major clusters and the minor clusters, calculating averages values of the centroid points in the clusters, and dividing the average values by the number of the plurality of image frames to detect the representative centroid point.

14. The characterizing point checking region setting method of claim 11, further comprising checking whether the characterizing point checking region includes a standard level of characterizing points of the image frames extracted for the certain period of time or greater.

19. An image stabilizing method comprising:

the characterizing point checking region setting method of claim 10;

setting the characterizing point checking region in an image frame that is currently input;

comparing the currently input image frame with a reference image that is preset; and

adjusting the currently input image frame as much as a shaking amount when it is determined that the current image frame is shaken.

20. The image stabilizing method of claim 19, further comprising:

extracting an image frame that is the least shaken among the plurality of image frames taken for a certain period of time; and

setting the extracted image frame as the reference image.