Dynamic video equalization of images using face-tracking

• US 7,706,576 B1
• Filed: 12/28/2004
• Issued: 04/27/2010
• Est. Priority Date: 12/28/2004
• Status: Active Grant

• Current Assignee: Avaya Incorporated
• Original Assignee: Avaya Incorporated
• Inventors: Beck, Alexander John Gray, Thambiratnam, David Preshan, Scholte, Alexander Martin, Choy, Jonathan R. Yee-Hang
• Primary Examiner(s): Mehta; Bhavesh M
• Assistant Examiner(s): Newman; Michael A
• Application Number:
  US11/025,493
• Time in Patent Office: 1,946 Days
• Field of Search: 382/118, 382/164, 382/173, 382/203, 382/209, 382/243, 382/274, 348/14.01, 348/14.12, 375/240.08
• US Class Current: 382/118
• CPC Class Codes: H04L 12/1827 Network arrangements for conference optimisation or adaptation
• CPC Class Codes: H04N 7/147 Communication arrangements, e.g. identifying the communication as a video-communication, intermediate storage of the signals selecting H04Q

Abstract

The present invention provides for the dynamic video equalization of images. Face tracking is used to identify a portion of an image corresponding to a human face. Those areas of the image identified as corresponding to a human face are optimized as compared to other areas of the image. Optimization is performed by allocating a greater number of image parameters to the area of the image corresponding to a human face than are allocated to other areas of the image. Accordingly, the portion of an image containing the human face is of higher quality as compared to other portions of the image.

Claims

1. A method for enhancing selected portions of an image, comprising:

a processor obtaining an image, wherein said image includes at least one human face;

the processor determining an area within said image corresponding to said at least one human face; and

the processor optimizing said image in said area corresponding to said at least one human face to obtain an optimized image;

wherein said optimizing said image in said area corresponding to said at least one human face includes;

allocating a greater number of available colors to a range of colors within said area and a smaller number of available colors to said image outside of said area;

allocating a greater number of available brightness levels to a brightness range within said area and a smaller number of available brightness levels to said image outside of said area; and

allocating a greater range of contrast to said at least one area and a smaller range of contrast to said image outside of said area;

wherein more than one color is included in said range of available colors allocated to portions of said image outside of said area corresponding to said human face,

wherein a greater portion of an available range of parameter values related to at least one of an available range of colors, available brightness levels, or range of contrast are devoted to said area corresponding to said at least one human face than to other areas of said image,

wherein a total amount of image information is constant across different frames of a sequence of images, and

wherein areas of said sequence of images corresponding to said at least one human face change across said different frames of said sequence of images.

2. The method of claim 1, wherein said image is represented by a finite number of parameters and parameter values, and wherein said optimizing said image in said area corresponding to said at least one human face includes allocating a greater number of said parameter values to image information included in said area corresponding to said at least one human face than to other areas of said image.

3. The method of claim 1, wherein said image includes a first number of human faces, wherein one or more areas corresponding to said first number of human faces is determined, and wherein said image is optimized in said one or more areas corresponding to said first number of human faces.

4. The method of claim 1, further comprising:

providing said optimized image to an output device.

5. The method of claim 4, further comprising:

initiating a video conference, wherein said image is obtained at a first location and wherein said output device is at a second location.

6. The method of claim 5, wherein said optimizing is performed by a video conferencing endpoint at said first location.

7. The method of claim 1, wherein said determining an area within said image corresponding to at least one human face comprises applying a segmentation algorithm.

8. The method of claim 1, wherein a quality of said area within said optimized image is superior to a quality of said optimized image outside of said area.

9. A system for providing dynamically equalized video output, comprising:

a first video input device;

a first video processor interconnected to said first video input device, wherein an area of an image obtained by said first video input device including a human face is optimized to obtain an optimized image, wherein a greater range of image parameters is allocated to said area including a human face than to other areas of said image, wherein a greater range of brightness levels is allocated to said area including a human face than to other areas of said image, wherein a greater range of contrast values is allocated to said area including a human face than to other areas of said image, wherein a plurality of image parameter values are used for said other areas of said image, wherein a greater portion of an available range of image parameter values are allocated to said area including a human face than to other areas of said image, wherein a total amount of available image information available for different frames in a sequence of images is constant, wherein said area of said sequence of images including a human face change for different frames in said sequence of images; and

a first video output device interconnected to said first video processor, wherein said optimized image is output by said first video output device.

10. The system of claim 9, further comprising:

a communication network, wherein said first video processor and said first video output device are interconnected by said communication network.

11. The system of claim 9, further comprising:

a communication network, wherein said first video input device and said first video processor are interconnected by said communication network.

12. The system of claim 9, further comprising a first video conferencing endpoint, wherein said first video input device is provided as part of said first video conferencing

endpoint, and wherein said first video processor is provided as part of said first video conferencing endpoint.

13. The system of claim 12, further comprising a second video conferencing endpoint, wherein said first video output device is provided as part of said second video conferencing endpoint.

14. The system of claim 13, wherein said second video conferencing endpoint further includes a second video input device, wherein said first video conferencing endpoint further includes a second video output device, and wherein said second video output device is operable to display image information from said second video input device.

15. The system of claim 14, wherein said second video conferencing endpoint further includes a second video processor interconnected to said second video input device, wherein an area of an image obtained by said second video input device including a human face is optimized to obtain an optimized image, and wherein said optimized image is displayed by said second video output device.

16. The system of claim 12, wherein said first video conferencing endpoint further includes an audio input device and an audio output device.

17. A system for dynamic video equalization of images, comprising:

means for collecting image data related to a first scene;

means for determining an area of said collected image data corresponding to an image of a human face; and

means for selectively optimizing said area of said collected image data corresponding to an image of a human face to obtain optimized image data, wherein said area of said collected image data corresponding to an image of a human face is optimized by the following;

allocating a greater number of available colors to a range of colors within said area than is allocated to portions of said collected image data outside of said area corresponding to an image of a human face;

allocating a greater number of available brightness levels to a brightness range within said area than is allocated to portions of said collected image data outside of said area corresponding to an image of a human face;

allocating a greater number of contrast levels to a contrast range within said area than is allocated to portions of said collected image data outside of said area, wherein more than one color is included in said range of available colors allocated to said portion outside of said area corresponding to an image of a human face, wherein a portion of the image data not corresponding to the image of a human face is represented by a range of values including a plurality of image parameter values, wherein a greater portion of image parameter values including at least one of a number of available colors, a number of available brightness levels, and a number of available contrast levels devoted to said area of collected image data corresponding to an image of a human face is greater than to other areas of said image, wherein a total amount of image information is constant across different frames of a sequence of images, and wherein said area of collected image data corresponding to an image of a human face changes across said different frames of said sequence of images.

18. The system of claim 17, further comprising:

means for displaying an optimized image from said optimized image data.

19. The system of claim 18, wherein said optimized image is displayed in substantially realtime.

20. The system of claim 18, further comprising:

means for providing a communication channel, wherein said optimized image data is transmitted to said means for displaying said optimized image.

21. The system of claim 17, wherein said means for selectively optimizing allocates a greater range of values associated with at least a first image parameter to said area of said collected image data than to another area of said collected image data.