Noise cleaning and interpolating sparsely populated color digital image using a variable noise cleaning kernel

Claim 1

A method for processing a sparsely populated color digital image provided by a color filter array of a scanner or digital camera having colored pixels to produce a fully populated and noise clean color image, comprising: noise cleaning the colored pixels in the sparsely populated color digital image by providing a noise cleaning kernel for each color in the sparsely populated image and calculating noise cleaned pixel values by averaging selected kernel pixel values based on the kernel pixel values and the pixel value of a pixel to be cleaned of the kernel respectively wherein each kernel operates upon a color pixel by changing values of neighboring pixels in accordance with noise thresholds, and wherein the averaging is accomplished by using a difference in value between a center pixel value and selected kernel pixel values; and interpolating the noise clean sparsely populated image for producing color pixels with appropriate values missing from the sparsely populated color digital image by interpolating the color values for missing pixels from neighboring color pixels.

Claim 2

A method for processing a sparsely populated color digital image provided by a color filter array of a scanner or digital camera having colored pixels to produce a fully populated and noise clean color image, comprising: noise cleaning the colored pixels in the sparsely populated color digital image by providing a noise cleaning kernel for each color in the sparsely populated image and calculating noise cleaned pixel values by averaging selected kernel pixel values based on the kernel pixel values and the pixel value of a pixel to be cleaned of the kernel respectively wherein each kernel operates upon a color pixel by changing values of neighboring pixels in accordance with noise thresholds, and wherein the averaging is accomplished by using a difference in value between a center pixel value and selected kernel pixel values; and interpolating the noise clean sparsely populated image for producing color pixels with appropriate values missing from the sparsely populated color digital image by interpolating the color values for missing pixels from neighboring color pixels.

Claim 3

A method for processing a sparsely populated color digital image provided by a color filter array of a sensor having colored pixels to produce a fully populated and noise clean color image, comprising: noise cleaning the colored pixels in the sparsely populated color digital image using a variably weighted noise cleaning kernel for each color in the sparsely populated image; and interpolating the noise cleaned sparsely populated image for producing color pixels with appropriate values missing from the sparsely populated color digital image by interpolating the color values for missing pixels from neighboring color pixels, and wherein a weighted contribution to a kernel calculation depends on an absolute value of a difference between kernel pixels values and a pixel value of the pixel to be cleaned.

Claim 4

A method for processing a sparsely populated color digital image provided by a color filter array of a sensor having colored pixels to produce a fully populated and noise clean color image, comprising: noise cleaning the colored pixels in the sparsely populated color digital image using a variably weighted noise cleaning kernel for each color in the sparsely populated image; and interpolating the noise cleaned sparsely populated image for producing color pixels with appropriate values missing from the sparsely populated color digital image by interpolating the color values for missing pixels from neighboring color pixels, and wherein a weighted contribution to a kernel calculation depends on an absolute value of a difference between kernel pixels values and a pixel value of the pixel to be cleaned.

Claim 5

A method for processing a sparsely populated color digital image provided by a color filter array of a sensor having colored pixels to produce a fully populated and noise clean color image, comprising: noise cleaning the colored pixels in the sparsely populated color digital image using a variably weighted noise cleaning kernel for each color in the sparsely populated image; and interpolating the noise cleaned sparsely populated image for producing color pixels with appropriate values missing from the sparsely populated color digital image by interpolating the color values for missing pixels from neighboring color pixels, wherein a pixel in the kernel is used in the cleaning responsive to a noise threshold for the pixel, wherein the each pixel value has a corresponding noise threshold; and wherein noise thresholds for the pixels of the kernel are derived from test images having spatially flat regions.

Claim 6

A method for processing a sparsely populated color digital image provided by a color filter array of a sensor having colored pixels to produce a fully populated and noise clean color image, comprising: noise cleaning the colored pixels in the sparsely populated color digital image using a variably weighted noise cleaning kernel for each color in the sparsely populated image; and interpolating the noise cleaned sparsely populated image for producing color pixels with appropriate values missing from the sparsely populated color digital image by interpolating the color values for missing pixels from neighboring color pixels, wherein a pixel in the kernel is used in the cleaning responsive to a noise threshold for the pixel, wherein the each pixel value has a corresponding noise threshold; and wherein noise thresholds for the pixels of the kernel are derived from test images having spatially flat regions.

Claim 7

A method for processing a sparsely populated color digital image provided by a color filter array of a sensor having colored pixels to produce a fully populated and noise clean color image, comprising: noise cleaning the colored pixels in the sparsely populated color digital image using a variably weighted noise cleaning kernel for each color in the sparsely populated image; and interpolating the noise cleaned sparsely populated image for producing color pixels with appropriate values missing from the sparsely populated color digital image by interpolating the color values for missing pixels from neighboring color pixels, wherein a pixel in the kernel is used in the cleaning responsive to a noise threshold for the pixel, and wherein each pixel has a corresponding noise threshold that is a function of the pixel value.

Claim 8

A method for processing a sparsely populated color digital image provided by a color filter array of a sensor having colored pixels to produce a fully populated and noise clean color image, comprising: noise cleaning the colored pixels in the sparsely populated color digital image using a variably weighted noise cleaning kernel for each color in the sparsely populated image; and interpolating the noise cleaned sparsely populated image for producing color pixels with appropriate values missing from the sparsely populated color digital image by interpolating the color values for missing pixels from neighboring color pixels, wherein a pixel in the kernel is used in the cleaning responsive to a noise threshold for the pixel, and wherein each pixel has a corresponding noise threshold that is a function of the pixel value.

Claim 10

A method for processing a sparsely populated color digital image having colored pixels to produce a fully populated and noise clean color image, comprising: noise cleaning the sparsely populated image providing a noise clean sparsely populated color digital image comprising: determining an absolute value of a difference between a pixel of the kernel and the pixel to be cleaned; comparing the absolute value to a corresponding noise threshold for a kernel pixel from a table of noise thresholds derived from a test image; and using the kernel pixel in the cleaning if the absolute value is less than or equal to the noise threshold; averaging kernel pixels in the cleaning; and interpolating the noise clean sparsely populated image producing a fully populated and noise clean color image.

Claim 11

A method for processing a sparsely populated color digital image having colored pixels to produce a fully populated and noise clean color image, comprising: noise cleaning the sparsely populated image providing a noise clean sparsely populated color digital image comprising: determining an absolute value of a difference between a pixel of the kernel and the pixel to be cleaned; comparing the absolute value to a corresponding noise threshold for a kernel pixel from a table of noise thresholds derived from a test image; and using the kernel pixel in the cleaning if the absolute value is less than or equal to the noise threshold; averaging kernel pixels in the cleaning; and interpolating the noise clean sparsely populated image producing a fully populated and noise clean color image.