Automatically Keying an Image

Abstract

Some embodiments provide a method for automatically selecting a portion of an image that includes several pixels, each of which has a set of pixel values. The method identifies a background color of the image. For each region of a set of regions of a color space that correspond to the background color, the method determines whether a threshold number of pixels in the image have pixel values that are in the region. The method identifies the pixels of the image in the regions that correspond to the background color and have a threshold number of pixels in the image. The method generates a portion of the color space to define a selection of the image using the pixel values of the identified pixels.

1 - 20 . (canceled) 
     
     
         21 . A non-transitory machine readable medium storing a program for execution by at least one processing unit, the program comprising sets of instructions for:
 automatically generating a key for an image comprising a plurality of pixels, wherein the key is an enclosed area of a color space for identifying a region of interest of the image;   automatically identifying a portion of the region of interest in the image having at least a threshold size, wherein a set of color values for each pixel in the portion is in the enclosed area of the color space; and   automatically generating a user-selectable item for display over the portion of the region of interest of the image, wherein modification of the user-selectable item modifies the key for the image.   
     
     
         22 . The non-transitory machine readable medium of  claim 21 , wherein the threshold size is measured in pixels. 
     
     
         23 . The non-transitory machine readable medium of  claim 21 , wherein the user-selectable item has a rectangular shape. 
     
     
         24 . The non-transitory machine readable medium of  claim 21 , wherein the set of instructions for automatically generating a key for the image comprises sets of instructions for:
 identifying a plurality of regions in the color space;   determining a count of pixels for each region of the plurality of regions;   identifying a set of regions of the plurality of regions, wherein the count of pixels for each region in the set of regions is greater than a threshold value; and   generating the key for the image based on the set of regions.   
     
     
         25 . The non-transitory machine readable medium of  claim 24 , wherein the threshold value is based on a maximum number of pixels in a single region of the plurality of regions. 
     
     
         26 . The non-transitory machine readable medium of  claim 21 , wherein the key for the image comprises a first sub-area of the color space and a second sub-area of the color space that encloses the first sub-area of the color space, wherein pixels with a set of color values in the first sub-area of the color space are fully selected and pixels with a set of color values in the second sub-area of the color space are partially selected. 
     
     
         27 . The non-transitory machine readable medium of  claim 21 , wherein the image is a video picture. 
     
     
         28 . The non-transitory machine readable medium of  claim 21 , wherein the color space is a first color space and the key is generated in a second color space that is a linear transformation of the first color space. 
     
     
         29 . A method for identifying a portion of a region of interest in an image, the method comprising:
 automatically generating a key for an image comprising a plurality of pixels, wherein the key is an enclosed area of a color space for identifying a region of interest of the image;   automatically identifying a portion of the region of interest in the image having at least a threshold size, wherein a set of values for each pixel in the portion is in the enclosed area of the color space; and   automatically generating a user-selectable item for display over the portion of the region of interest of the image, wherein modification of the user-selectable item modifies the key for the image.   
     
     
         30 . The method of  claim 29 , wherein the threshold size is measured in pixels. 
     
     
         31 . The method of  claim 29 , wherein the user-selectable item has a rectangular shape. 
     
     
         32 . The method of  claim 29 , wherein automatically generating a key for the image comprises:
 identifying a plurality of regions in the color space;   determining a count of pixels for each region of the plurality of regions;   identifying a set of regions of the plurality of regions, wherein the count of pixels for each region in the set of regions is greater than a threshold value; and   generating the key for the image based on the set of regions.   
     
     
         33 . The method of  claim 32 , wherein the threshold value is based on a maximum number of pixels in a single region of the plurality of regions. 
     
     
         34 . The method of  claim 29 , wherein the key for the image comprises a first sub-area of the color space and a second sub-area of the color space that encloses the first sub-area of the color space, wherein pixels with a set of color values in the first sub-area of the color space are fully selected and pixels with a set of color values in the second sub-area of the color space are partially selected. 
     
     
         35 . The method of  claim 29 , wherein the image is a video picture. 
     
     
         36 . The method of  claim 29 , wherein the color space is a first color space and the key is generated in a second color space that is a linear transformation of the first color space. 
     
     
         37 . An image editing system comprising:
 a key generation module for automatically generating a key for an image comprising a plurality of pixels, wherein the key is an enclosed area of a color space for identifying a region of interest of the image;   an identification module for automatically identifying a portion of the region of interest in the image having at least a threshold size, wherein a set of color values for each pixel in the portion is in the enclosed area of the color space; and   an interface module for automatically generating a user-selectable item for display over the portion of the region of interest of the image, wherein modification of the user-selectable item modifies the key for the image.   
     
     
         38 . The system of  claim 37 , wherein the user-selectable item has a rectangular shape. 
     
     
         39 . The system of  claim 37 , wherein the key generation module generates the key for the image by:
 identifying a plurality of regions in the color space;   determining a count of pixels for each region of the plurality of regions;   identifying a set of regions of the plurality of regions, wherein the count of pixels for each region in the set of regions is greater than a threshold value; and   generating the key for the image based on the set of regions.   
     
     
         40 . The system of  claim 37 , wherein the key for the image comprises a first sub-area of the color space and a second sub-area of the color space that encloses the first sub-area of the color space, wherein pixels with a set of color values in the first sub-area of the color space are fully selected and pixels with a set of color values in the second sub-area of the color space are partially selected.