US 6,633,670 B1 · Mask generation for multi-layer image decomposition

Abstract

To improve compression efficiency and decompressed image quality, images are decomposed into separate layers each containing a limited number of image element types (text, line art, or photographic). Each layer can then be compressed separately with a process that is optimal for the type of image element included in the layer. Images are decomposed into foreground, background, and mask layers. The value of a pixel in the mask layer designates whether the spatially corresponding pixel of the recomposed image will be selected from the spatially corresponding pixels of either the foreground or the background layer. The present invention provides a method of determining the value of a pixel for a mask layer. The image is partitioned into large and small sub-images or blocks. A sub-image mask is created for each sub-image by sorting pixels of the sub-image into clusters centered on the luminance of pixels of a pair of pixels of maximum luminance gradient. The value of a pixel in the mask layer is determined by a relationship of the sub-image mask values of the pixels larger and smaller sub-images of which the pixel being evaluated is a member.

Claims26

Dependent claims

A method of generating a mask for a layered decomposition of an image, said method comprising the steps of:

(a) partitioning said image into a plurality of first sub-images and into a plurality of second sub-images dimensionally differing from said first sub-images, said first and said second sub-images comprising pluralities of pixels;
(b) assigning a first sub-image mask value to an evaluation pixel according to a relationship of a luminance of said evaluation pixel and a sub-image luminance of a first sub-image containing said evaluation pixel;
(c) assigning a second sub-image mask value to said evaluation pixel according to a relationship of said luminance of said evaluation pixel and a sub-image luminance of a second sub-image containing said evaluation pixel;
and
(d) setting a mask value for a pixel in said mask spatially corresponding to said evaluation pixel according to a relationship of said first and second sub-image mask values for a plurality of pixels of said first and said second sub-images containing said evaluation pixel.

2 more elements

A method of generating a mask for a layered decomposition of an image, the method comprising the steps of:

(a) determining a gradient relating a pixel of said image to a neighboring pixel;
(b) spatially partitioning said image into a plurality of first sub-images each comprising a first plurality of pixels;
(c) assigning a first sub-image mask value to an evaluation pixel according to a relationship of a luminance of said evaluation pixel and a sub-image luminance of said first sub-image containing said evaluation pixel;
(d) spatially partitioning said image into a plurality of second sub-images each comprising a second plurality of pixels;
(e) assigning a second sub-image mask value to said evaluation pixel according to a relationship of said luminance of said evaluation pixel and a sub-image luminance of said second sub-image containing said evaluation pixel;
and
(f) setting a mask value for said evaluation pixel as a function of said first and second sub-image mask values for a plurality of pixels of said first and said second sub-images containing said evaluation pixel.

4 more elements

A method of layered decomposition of an image, said method comprising the steps of:

(a) determining a luminance of a plurality of pixels of said image;
(b) calculating a gradient between said luminance of a pixel and said luminance of a horizontally adjacent pixel and a vertically adjacent pixel;
(c) partitioning said image into a first plurality of mutually exclusive first sub-images each comprising a first plurality of pixels and a second plurality of mutually exclusive second sub-images each comprising a second plurality of pixels;
(d) identifying a pair of pixels of maximum gradient for a sub-image,
(e) associating an evaluation pixel with a luminously nearer one of a first pixel of a maximum gradient pair and a second pixel of a maximum gradient pair of pixels of a sub-image that contains said evaluation pixel and said maximum gradient pair of pixels;
(f) assigning a sub-image mask value to a pixel associated with a pixel of said maximum gradient pair of pixels according to a luminance relationship of said pixels of said maximum gradient pair;
(g) assigning a flat sub-image mask value to pixels of a sub-image if said maximum gradient is less than a gradient threshold;
(h) assigning a mask layer value to a pixel as a function of a sub-image mask value assigned to a plurality of pixels of said sub-image containing said pixel;
and
(i) selecting a spatially corresponding pixel for inclusion in an image from one of a background layer and foreground layer in response to said mask layer value of said pixel.

7 more elements

Dependencies are inferred from the claim text.

Mask generation for multi-layer image decomposition

Life of this patent

Mask generation for multi-layer image decomposition

Life of this patent

Abstract

Claims26