Automatic tone adjustment by contrast gain-control on edges

Claim 1

A method of reducing the dynamic range of an input image for effectively rendering the input image on an output display medium, the method comprising the steps of: (a) extracting detail signals and coarse signals from the input image; (b) generating contrast gain-control signals from the input image by detecting coarse scale edges; (c) modifying the detail signals according to the contrast gain signals; and, (d) adding the modified detail signals to the coarse signals to obtain an output image.

Claim 2

A method of reducing the dynamic range of an input image for effectively rendering the input image on an output display medium, the method comprising the steps of: (a) extracting detail signals and coarse signals from the input image; (b) generating contrast gain-control signals from the input image by detecting coarse scale edges; (c) modifying the detail signals according to the contrast gain signals; and, (d) adding the modified detail signals to the coarse signals to obtain an output image.

Claim 6

A method of reducing the dynamic range of an input image I so that it can be rendered effectively on an output display medium, comprising the steps of: (a) constructing a tone scale curve T from the input image I; (b) mapping the input image I through the tone scale curve T to produce a tone-scaled image I′=T(I); (c) constructing a mask image of weighting factors, W, from the input image such that W varies from 1.0 in regions where edge contrast is low to 0.0 in regions where edge contrast is high; and, (d) producing an output image, I out =W H(I)+(1−W) H(I′)+L(I′), where H(I) is the high-pass filtered input image, H(I′) is the high-pass filtered tone-scaled image, and L(I′) is the low-pass filtered version of the tone-scaled image I′.

Claim 7

A method of reducing the dynamic range of an input image I so that it can be rendered effectively on an output display medium, comprising the steps of: (a) constructing a tone scale curve T from the input image I; (b) mapping the input image I through the tone scale curve T to produce a tone-scaled image I′=T(I); (c) constructing a mask image of weighting factors, W, from the input image such that W varies from 1.0 in regions where edge contrast is low to 0.0 in regions where edge contrast is high; and, (d) producing an output image, I out =W H(I)+(1−W) H(I′)+L(I′), where H(I) is the high-pass filtered input image, H(I′) is the high-pass filtered tone-scaled image, and L(I′) is the low-pass filtered version of the tone-scaled image I′.

Claim 10

A method of reducing the dynamic range of an input image I so that it can be rendered effectively on an output display medium, comprising the steps of: (a) constructing a tone scale curve T from the input image I; (b) mapping the input image I through a tone scale curve T to produce a tone-scaled image I′=T(I); (c) producing a difference image D=I−I′ by subtracting the tone-scaled image from the input image; (d) producing a modified difference image, D′, by processing the difference image D through a plurality of band-pass and high-pass filters; and, (e) adding the modified difference image D′ to the tone-scaled image I′ to produce a dynamic range compressed output image.

Claim 11

A method of reducing the dynamic range of an input image I so that it can be rendered effectively on an output display medium, comprising the steps of: (a) constructing a tone scale curve T from the input image I; (b) mapping the input image I through a tone scale curve T to produce a tone-scaled image I′=T(I); (c) producing a difference image D=I−I′ by subtracting the tone-scaled image from the input image; (d) producing a modified difference image, D′, by processing the difference image D through a plurality of band-pass and high-pass filters; and, (e) adding the modified difference image D′ to the tone-scaled image I′ to produce a dynamic range compressed output image.

Claim 14

A method of reducing the dynamic range of an input image I so that it can be rendered effectively on an output display medium, comprising the steps of: (a) constructing a tone scale curve T from the input image I; (b) mapping the input image I through the tone scale curve T to produce a tone-scaled image I′=T(I); (c) producing a difference image D=I−I′ by subtracting the tone-scaled image from the input image; (d) generating contrast gain-control signals G from the input image; (e) producing a modified difference image, D′, by passing the difference image D through an edge-detecting filter bank where edge signals in each spatial-scale are modulated by the said contrast gain-control signals G; and, (f) adding the modified difference image D′ to the tone-scaled image I′ to produce a dynamic range compressed output image.

Claim 15

A method of reducing the dynamic range of an input image I so that it can be rendered effectively on an output display medium, comprising the steps of: (a) constructing a tone scale curve T from the input image I; (b) mapping the input image I through the tone scale curve T to produce a tone-scaled image I′=T(I); (c) producing a difference image D=I−I′ by subtracting the tone-scaled image from the input image; (d) generating contrast gain-control signals G from the input image; (e) producing a modified difference image, D′, by passing the difference image D through an edge-detecting filter bank where edge signals in each spatial-scale are modulated by the said contrast gain-control signals G; and, (f) adding the modified difference image D′ to the tone-scaled image I′ to produce a dynamic range compressed output image.

Claim 18

A method of generating a tone-scale curve from an image for dynamic range compression, comprising the steps of: (a) computing a histogram from the input image; (b) finding a low end, a median point, and a high end of the input values from the histogram at some fixed percentiles; (c) mapping the low end point to a black level of the output medium with a higher than normal local slope; (d) mapping the median point to a medium gray level of the output medium with a normal contrast; and, (e) mapping the high end point to a white level of the output medium with an adjustable contrast, depending on the shape and the range of an upper tail of the histogram.

Claim 19

A method of generating a tone-scale curve from an image for dynamic range compression, comprising the steps of: (a) computing a histogram from the input image; (b) finding a low end, a median point, and a high end of the input values from the histogram at some fixed percentiles; (c) mapping the low end point to a black level of the output medium with a higher than normal local slope; (d) mapping the median point to a medium gray level of the output medium with a normal contrast; and, (e) mapping the high end point to a white level of the output medium with an adjustable contrast, depending on the shape and the range of an upper tail of the histogram.

Claim 20

A method of image contrast enhancement using contrast gain-control signals for suppressing banding artifacts, comprising the steps of: (a) decomposing an input image into edge gradients of several spatial scales; (b) generating contrast gain-control signals G from the original input image; (c) enhancing each edge gradient signal by modifying its amplitude according to separately generated contrast gain-control signals G in step (b); and, (d) reconstructing an enhanced output image from the modified edge signals of step (c).

Claim 21

A method of image contrast enhancement using contrast gain-control signals for suppressing banding artifacts, comprising the steps of: (a) decomposing an input image into edge gradients of several spatial scales; (b) generating contrast gain-control signals G from the original input image; (c) enhancing each edge gradient signal by modifying its amplitude according to separately generated contrast gain-control signals G in step (b); and, (d) reconstructing an enhanced output image from the modified edge signals of step (c).