Activity detector

• US 7,263,227 B2
• Filed: 04/25/2002
• Issued: 08/28/2007
• Est. Priority Date: 04/25/2002
• Status: Expired due to Fees

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Simard, Patrice Y.
• Primary Examiner(s): Mancuso; Joseph
• Assistant Examiner(s): LIEW, ALEX KOK SOON
• Application Number:
  US10/133,842
• Publication Number:
  US 20030202696A1
• Time in Patent Office: 1,951 Days
• Field of Search: None
• US Class Current: 382/195
• CPC Class Codes: G06V 30/413 Classification of content, e.g. text, photographs or tables

Abstract

A system and method facilitating activity (e.g., dithering/half toning and/or noise) detection is provided. The invention includes an activity detection system having a connected component analyzer and an activity detector. The invention provides for the quantity of connected component(s) in and/or intersecting a region surrounding a pixel to be determined. The activity detector provides an activity map output based, at least in part, upon the quantity of connected component(s) in and/or intersecting the region. The invention further provides for an optional image processor. In one example, if the quantity exceeds a first threshold, dithering/half toning is detected and appropriate action can be taken. Additionally, if the quantity is less than a second threshold, noise is detected and appropriate action can be taken.

Claims

1. An activity detection system, comprising:

a connected component analyzer that determines a quantity of connected components in or intersecting a region surrounding a pixel by maintaining an identity of substantially all the quantity of connected components and a relative quantity of each connected component ion the region surrounding the pixel;

an activity detector that provides an activity map output based, at least in part, upon the quantity of connected components in or intersecting the region;

wherein said activity detector identifies bounding boxes of the connected components, said bounding boxes are enlarged and creates a derivate image of the enlarged bounding boxes; and

an image processor that assigns the pixel to one of a background or a foreground based at least in part upon the quantity of connected components in or intersecting the region surrounding the pixel;

wherein the image processor assigns to the foreground based at least in part upon a determination that the quantity of connected components exceeds a first threshold, and assigns the pixel to the background based at least in part upon a determination that the quantity of connected components is less than a second threshold.

2. The activity detection system of claim 1, the image processor determines a quantity of pixels in the background of the region and a quantity of pixels in the foreground of the region, the activity detector assigns the pixel to the foreground if the quantity of pixels in the foreground of the region is greater than the quantity of pixels in the background of the region, the activity detector otherwise assigns the pixel to the background.

3. The activity detection system of claim 1, the image processor assigns the pixel to the background if the quantity of connected components is less than a third threshold.

4. The activity detection system of claim 1, the connected component analyzer and the image processor receives a bi-level image input comprising a plurality of pixels assigned to at least one of the foreground and the background, the image processor provides a bi-level image output, the bi-level image output being based at least in part upon the bi-level image input and assignment of at least one pixel to one of the background or the foreground based at least in part upon the activity map output.

5. The activity detection system of claim 1, the activity detector further identifies bounding boxes of the connected components.

6. The activity detection system of claim 5, the activity detector further enlarges the bounding boxes.

7. The activity detection system of claim 6, the activity detector further creates a derivate image of the enlarged bounding boxes.

8. The activity detection system of claim 7, the activity detector provides an output associated with a double integral of the derivate image.

9. The activity detection system of claim 6, the activity detector further creates a derivate image of a lower resolution of the enlarged bounding boxes.

10. The activity detection system of claim 9, the activity detector provides an output associated with a double integral of the derivate image.

11. The activity detection system of claim 1, the activity detector further dilates the activity map output.

12. The activity detection system of claim 1, the activity detector further erodes the activity map output.

13. The activity detection system of claim 1, the activity detector further smoothes the activity map output.

14. A photocopier employing the activity detection system of claim 1.

15. A document scanner employing the activity detection system of claim 1.

16. An optical character recognition system employing the activity detection system of claim 1.

17. A personal digital assistant employing the activity detection system of claim 1.

18. A fax machine employing the activity detection system of claim 1.

19. A digital camera employing the activity detection system of claim 1.

20. A digital video camera employing the activity detection system of claim 1.

21. A segmented layered image system employing the activity detection system of claim 1.

22. A tablet personal computer employing the activity detection system of claim 1.

23. A method for detecting activity in an image, comprising:

identifying bounding boxes of connected components;

enlarging the bounding boxes;

creating a derivative image based, at least in part upon the enlarged bounding boxes, the creation of the derivative image includes maintaining an identity of substantially all connected components circumscribed by the enlarged bounding boxes and a relative quantity associated with each connected component in a region surrounding a pixel;

providing an output based, at least in part, upon a double integral of the derivate image; and

assigning pixels to a foreground if a number of connected components associated therewith exceeds a first threshold level and to a background if the number of connected components in and intersecting the region surrounding the pixel is less than or equal to a second threshold level.

24. A method for detecting activity in an image, comprising:

calculating a quantity of connected components in or intersecting a region surrounding a pixel, the calculation includes at least maintaining an identity of connected components and a relative quantity associated with each connected component in the region, the relative quantity determined with respect to the pixel;

wherein identifying bounding boxes of the connected components, wherein said bounding boxes are enlarged and creating a derivate image of the enlarged bounding boxes; and

assigning the pixel to one of a background or a foreground based at least in part upon the quantity of connected components in or intersecting the region surrounding the pixel;

wherein the pixel is assigned to the foreground if the quantity of connected components in or intersecting the region surrounding the pixel is greater than a first threshold and to the background if the quantity of connected components in or intersecting the region surrounding the pixel is less than or equal to a second threshold.

25. The method of claim 24, further comprising identifying connected components in or intersecting the region surrounding the pixel.

26. The method of claim 24, further comprising at least one of the following:

determining whether the quantity of connected components in or intersecting the region exceeds a first threshold;

determining a quantity of pixels in the background of the region and a quantity of pixels in the foreground of the region; and

,

assigning the pixel to the foreground if the quantity of pixels in the foreground of the region is greater than the quantity of pixels in the background of the region.

27. The method of claim 24, further comprising assigning the pixel to the background if the quantity of connected components is less than a third threshold.

28. A method for detecting dithering in an image, comprising:

calculating a quantity of connected components in or intersecting a region surrounding a pixel by establishing a relative quantity associated with each connected component in the region from the perspective of the pixel and maintaining an identity of all connected components in the quantity of connected components;

wherein identifying bounding boxes of the connected components, wherein said bounding boxes are enlarged and creating a derivate image of the enlarged bounding boxes;

determining whether the quantity of connected component in or intersecting the region exceeds a first threshold; and

assigning the pixel to a foreground, if the quantity of connected component in or intersecting the region surrounding the pixel is greater than a first threshold, and to a background if the quantity of connected components is less than or equal to a second threshold.

29. The method of claim 28, further comprising at least one of the following:

determining a quantity of pixels in the background of the region and a quantity of pixels in the foreground of the region; and

,

assigning the pixel to the foreground if the quantity of pixels in the foreground of the region is greater than the quantity of pixels in the background of the region.

30. A method for detecting noise in an image, comprising:

calculating a quantity of connected components surrounding a pixel in a region, wherein identifying bounding boxes of the connected components, wherein said bounding boxes are enlarged and creating a derivate image of the enlarged bounding boxes; and

assigning the pixel to one of the background or the foreground based at least in part upon a determination that the quantity of connected components is less than a second threshold or greater than a first threshold, the quantity of connected components established via maintenance of substantially all connected components and utilization of a relative quantity of each connected component in the region, the relative quantity establish with respect to the pixel.

31. A computer readable medium storing computer executable components of a system for activity detection, comprising:

a connected component analyzer component that determines a quantity of connected components in or intersecting a region surrounding a pixel by maintaining an identity of substantially all connected components and a relative quantity associated with each connected component include in the quantity of connected components, the relative quantity ascertained with respect to the pixel, wherein identifying bounding boxes of the connected components, wherein said bounding boxes are enlarged and creating a derivate image of the enlarged bounding boxes; and

an activity detector component that provides an activity output based, at least in part, upon the quantity of connected components in or intersecting the region, wherein the pixel is assigned to a foreground of the image if the quantity of connected components in or intersecting a region around the pixel is greater than a first threshold and to a background if the quantity of connected components in or intersecting a region around the pixel is less than or equal to a second threshold.

32. A computer readable medium storing computer executable instructions operable to perform a method for activity detection, comprising:

identifying bounding boxes of connected components;

enlarging the bounding boxes;

creating a derivative image based, at least in part upon the enlarged bounding boxes, the derivative image created at least by maintaining an identity of all connected components included in the enlarged bounding boxes and establishing a relative quantity associated with each connected component in a region surrounding a pixel;

providing an output based, at least in part, upon a double integral of the derivate image; and

analyzing the output image to assign a pixel therein to one of a foreground and a background based at least in part on whether a number of connected components in and/or intersecting a region about the pixel is grater than a first threshold or less than or equal to a second threshold, respectively.

33. A computer readable medium storing computer executable instructions operable to perform a method for activity detection, comprising:

calculating a quantity of connected components in or intersecting a region surrounding a pixel by maintaining an identity of at least all the quantity of connected components and establishing a relative quantity associated with each connected component included in the quantity of connected components in the region surrounding the pixel, wherein identifying bounding boxes of the connected components, wherein said bounding boxes are enlarged and creating a derivate image of the enlarged bounding boxes; and

assigning the pixel to one of a background or a foreground based at least in part upon the quantity of connected components in or intersecting the region surrounding the pixel, wherein the pixel is assigned to the foreground if the quantity of connected components exceeds a first threshold and to the background if the quantity is less than a second threshold.

34. An activity detector, comprising:

means for determining a quantity of connected components in or intersecting a region surrounding a pixel;

means for maintaining an identity of at least all the quantity of connected components and establishing a relative quantity associated with each connected component that comprises the identity in the region, the relative quantity determined from the perspective of the pixel, wherein identifying bounding boxes of the connected components, wherein said bounding boxes are enlarged and creating a derivate image of the enlarged bounding boxes; and

,

means for assigning the pixel to one of a background or a foreground based at least in part upon the quantity of connected components in or intersecting the region surrounding the pixel, wherein the means for assigning assigns the pixel to a foreground if the quantity is equal to or greater than a first threshold and assigns the pixel to the background if the quantity is less than or equal to a second threshold.