Granted Patent B2

US 9,165,192 · App. 13/206,147 · Granted Oct 20, 2015

Abstract

Claims (30)

1. An apparatus for separating an image into a foreground and a background, the apparatus comprising:

an edge image generating unit which generates an edge image of a current image, wherein the current image includes a background and a foreground;

a background edge model renewing unit which renews a background edge model of a previous image based on the generated edge image of the current image; and

a foreground edge extracting unit which generates a foreground edge image based on the generated edge image and the renewed background edge model,

wherein if a pixel is an edge pixel of the edge image of the current image, the background edge model renewing unit increases a pixel value of a corresponding pixel of the background edge model of the previous image by a first predetermined value, and

wherein if the pixel is a non-edge pixel of the edge image of the current image, the background edge model renewing unit decreases the pixel value of the corresponding pixel of the background edge model of the previous image by a second predetermined value.

2. The apparatus of claim 1 , wherein the pixel value of the background edge model is between a maximum value and a minimum value.

3. The apparatus of claim 1 , wherein if a pixel of the background edge model corresponding to an edge pixel of the edge image has a pixel value larger than a first threshold value, the foreground edge extracting unit determines the pixel as being from a background edge, and

wherein if the pixel value of the pixel is smaller than the first threshold value, the foreground edge extracting unit determines the pixel as being from a foreground edge.

4. The apparatus of claim 3 , further comprising a foreground edge correction unit which deletes a foreground edge having a length shorter than a second threshold value.

5. The apparatus of claim 1 , wherein the edge image is a canny edge image.

6. The apparatus of claim 1 , further comprising an initial background edge model generating unit which generates an initialized background edge model.

7. A method of separating an image into a foreground and a background, the method comprising:

generating an edge image of a current image including the background and the foreground;

renewing a background edge model of a previous image based on the generated edge image of the current image; and

generating a foreground edge image based on the generated edge image and the renewed background edge model,

wherein the renewing of the background edge model of the previous image comprises increasing, by a first predetermined value, a pixel value of a pixel of the background edge model of the previous image if the pixel of the background edge model of the previous image corresponds to an edge pixel of the edge image of the current image, and decreasing, by a second predetermined value, the pixel value of the pixel of the background edge model of the previous image if the pixel of the background edge model of the previous image corresponds to a non-edge pixel of the edge image of the current image.

8. The method of claim 7 , wherein the pixel value of the background edge model is between a maximum value and a minimum value.

9. The method of claim 7 , wherein the generating of the foreground edge image comprises determining whether a pixel of the background edge model corresponding to an edge pixel of the edge image has a pixel value larger than a first threshold value, and if the pixel value is larger than the first threshold value, the pixel is determined as being from a background edge and if the pixel value is smaller than the first threshold value, the pixel is determined as being from a foreground edge.

10. The method of claim 9 , further comprising deleting a foreground edge having a length shorter than a second threshold value.

11. The method of claim 7 , wherein the edge image is a canny edge image.

12. The method of claim 7 , further comprising generating an initialized background edge model before generating the edge image.

13. An intelligent monitoring system comprising:

an image input unit which receives a current image corresponding to a surveillance area;

an image analyzing unit which processes the current image received by the image input unit and senses an event occurring within the surveillance area;

a storage unit which stores information about the surveillance area based on sensing of the event by the image analyzing unit; and

a display unit which displays a surveillance image based on the information stored in the storage unit,

wherein the image analyzing unit processes the current image received by the image input unit by generating an edge image of the current image, the edge image including a background and a foreground, renewing a background edge model of a previous image based on the generated edge image of the current age, and generating a foreground edge image based on the generated edge image and the renewed background edge model,

wherein the image analyzing unit senses the event occurring within the surveillance area based on the generated foreground edge image, and

wherein the renewing of the background edge model of the previous image comprises increasing, by a first predetermined value, a pixel value of a pixel of the background edge model of the previous image if the pixel of the background edge model of the previous image corresponds to an edge pixel of the edge image of the current image, and decreasing, by a second predetermined value, the pixel value of the pixel if the pixel of the background edge model of the previous image corresponds to a non-edge pixel of the edge image of the current image.

Assignments (6)