Method and apparatus for converting a halftone image into a continuous tone image

Claim 1

A method for converting a binary image into a gray scale image comprising: (a) selecting a neighborhood of pixels in said binary image, each pixel having a pixel value (X old ) at one of two levels, said neighborhood including a selected pixel; (b) determining a first value (μ), said step of determining said first value including computing an average value of said pixel values of said pixels within said neighborhood; (c) determining a second value (v), said step of determining said second value including selecting a numerical measure which represents a local variation of said pixel values among said pixels within said neighborhood, said step of determining said second value further including calculating said numerical measure; (d) adjusting said second value (v) by a weighting (γ), including multiplying said second value (v) by said weighting (γ), resulting in a weighted variation (γv); (e) determining an adjusted pixel value (x new ) of said pixel value of said selected pixel, wherein ##EQU5## (f) repeating steps (a)-(e) for other pixels of said binary image such that a gray scale image is formed; and (g) low pass filtering said gray scale image.

Claim 2

A method for converting a binary image into a gray scale image comprising: (a) selecting a neighborhood of pixels in said binary image, each pixel having a pixel value (X old ) at one of two levels, said neighborhood including a selected pixel; (b) determining a first value (μ), said step of determining said first value including computing an average value of said pixel values of said pixels within said neighborhood; (c) determining a second value (v), said step of determining said second value including selecting a numerical measure which represents a local variation of said pixel values among said pixels within said neighborhood, said step of determining said second value further including calculating said numerical measure; (d) adjusting said second value (v) by a weighting (γ), including multiplying said second value (v) by said weighting (γ), resulting in a weighted variation (γv); (e) determining an adjusted pixel value (x new ) of said pixel value of said selected pixel, wherein ##EQU5## (f) repeating steps (a)-(e) for other pixels of said binary image such that a gray scale image is formed; and (g) low pass filtering said gray scale image.

Claim 8

A method for converting a binary image into a gray scale image comprising: providing a binary image in the form of a first two-dimensional element array of binary image elements, each binary image element in said element array having an element value; constructing a first gray scale image in the form of a second two-dimensional element array of gray scale elements, said gray scale elements corresponding to said binary image elements in a one-to-one relationship, each gray scale element in said second element array having an element value, said step of constructing including; (a) for each binary image element in said first two-dimensional element array, determining a first value representing an averaging of element values for binary image elements in a window containing said each binary image element; (b) determining a second value representing element value variations within said window; (c) assigning a third value to the gray scale element in said second array corresponding to said each binary image element, said assigning including selecting a third value that is between said element value of said each binary image element and said first value only if the difference between said element value and said first value exceeds a selected weighted multiple of said second value, said assigning including selecting a third value that is equal to said element value of said each binary image element only if the difference between said element value of said each binary image element and said first value is one of equal to and less than said weighted multiple of said second value; low pass filtering said first gray scale image; and constructing a second gray scale image from said first gray scale image by following said sequence of steps for constructing said first gray scale image, the weighted multiple for constructing said second gray scale image being different than said weighted multiple for constructing said first gray scale image.

Claim 9

A method for converting a binary image into a gray scale image comprising: providing a binary image in the form of a first two-dimensional element array of binary image elements, each binary image element in said element array having an element value; constructing a first gray scale image in the form of a second two-dimensional element array of gray scale elements, said gray scale elements corresponding to said binary image elements in a one-to-one relationship, each gray scale element in said second element array having an element value, said step of constructing including; (a) for each binary image element in said first two-dimensional element array, determining a first value representing an averaging of element values for binary image elements in a window containing said each binary image element; (b) determining a second value representing element value variations within said window; (c) assigning a third value to the gray scale element in said second array corresponding to said each binary image element, said assigning including selecting a third value that is between said element value of said each binary image element and said first value only if the difference between said element value and said first value exceeds a selected weighted multiple of said second value, said assigning including selecting a third value that is equal to said element value of said each binary image element only if the difference between said element value of said each binary image element and said first value is one of equal to and less than said weighted multiple of said second value; low pass filtering said first gray scale image; and constructing a second gray scale image from said first gray scale image by following said sequence of steps for constructing said first gray scale image, the weighted multiple for constructing said second gray scale image being different than said weighted multiple for constructing said first gray scale image.

Claim 14

An apparatus for converting a binary image into a gray scale image comprising: input means for providing a binary image; an image processing pipeline having a plurality of stages, at least some of said stages including a low pass filter and a non-linear statistical smoothing means for determining on a pixel-by-pixel basis desired pixel values for a conversion of said binary image to a gray scale image, wherein for each pixel location (m,n) said desired pixel value (x m ,n new ) for said gray scale image is determined from the pixel value (x m ,n old ) of the input of said stage in accordance with the condition that ##EQU9## where μ m ,n is an average of pixel values within a pixel neighborhood including said pixel m,n and where γv m ,n is a weighted characterization of pixel value variations within said pixel neighborhood; and means for achieving one of visual presentation and image processing of said gray scale image converted by said image processing pipeline.

Claim 15

An apparatus for converting a binary image into a gray scale image comprising: input means for providing a binary image; an image processing pipeline having a plurality of stages, at least some of said stages including a low pass filter and a non-linear statistical smoothing means for determining on a pixel-by-pixel basis desired pixel values for a conversion of said binary image to a gray scale image, wherein for each pixel location (m,n) said desired pixel value (x m ,n new ) for said gray scale image is determined from the pixel value (x m ,n old ) of the input of said stage in accordance with the condition that ##EQU9## where μ m ,n is an average of pixel values within a pixel neighborhood including said pixel m,n and where γv m ,n is a weighted characterization of pixel value variations within said pixel neighborhood; and means for achieving one of visual presentation and image processing of said gray scale image converted by said image processing pipeline.