Simple and robust color saturation adjustment for digital images
A method and system for adjusting saturation in digital images that operates as closely as possible to the long-, medium-, short-(LMS) cone spectral response space. According to the method, a sensor component image such as an RGB image from a digital imager is input and converted to the LMS space. White point adaptation and equalization are performed on the LMS data. The saturation adjustment is then performed by applying a stretching transformation to the L and S LMS components with respect to the M component of each pixel.
1. A method of adjusting color saturation of a first color image comprising components of a first color space, said method being executed by an image processor and comprising:
creating a color space transformation matrix;
transforming the first color image into a second image using the transformation matrix, the second image having a plurality of long-wavelength (L), medium-wavelength (M), and short-wavelength (S) components;
determining white point adaptation levels for each of the plurality of long-wavelength (L), medium-wavelength (M), and short-wavelength (S) components by the image processor;
adjusting each of the plurality of long-wavelength (L), medium-wavelength (M), and short- wavelength (S) components by applying the associated white point adaptation levels to the plurality of long-wavelength (L), medium-wavelength (M), and short-wavelength (S) components;
stretching the adjusted long-wavelength (L) components and the short-wavelength (S) components with respect to the adjusted medium-wavelength (M) component;
applying the stretched L components and S components to the second image; and
transforming the second image into a new image having components of the first color space, wherein each L component is stretched according to the equation L 1 =M+K(L 0 −M), and where L 0 is an initial long-wavelength component, L 1 is a stretched long-wavelength component and K is a saturation adjustment factor.
2. The method of claim 1 , wherein the first color image is an RGB image.
3. A method of adjusting color saturation of a first color image comprising components of a first color space, said method being executed by an image processor and comprising:
creating a color space transformation matrix;
transforming the first color image into a second image using the transformation matrix, the second image having a plurality of long-wavelength (L), medium-wavelength (M), and short- wavelength (S) components;
determining white point adaptation levels for each of the plurality of long-wavelength (L), medium-wavelength (M), and short-wavelength (S) components by the image processor;
adjusting each of the plurality of long-wavelength (L), medium-wavelength (M), and short- wavelength (S) components by applying the associated white point adaptation levels to the plurality of long-wavelength (L), medium-wavelength (M), and short-wavelength (S) components;
stretching the adjusted long-wavelength (L) components and the short-wavelength (S) components with respect to the adjusted medium-wavelength (M) component, applying the stretched L components and S components to the second image; and
transforming the second image into a new image having components of the first color space, wherein each S component is stretched according to the equation S 1 =M+K(S 0 −M), and where K is a saturation adjustment factor, SO is an initial short-wavelength component, and S 1 is a stretched short-wavelength component.
4. The method of claim 3 , wherein each L component is stretched according to the equation L 1 =M+K(L 0 −M), where L 0 is an initial long-wavelength component, and L 1 is a stretched long-wavelength component.
5. An imaging device comprising:
an array of pixels;
readout circuitry coupled to the array, the circuitry adapted to output a first color image of a first color space based on signals received from the array; and
an image processor adapted to input the first color image and perform saturation adjustment on the first color image, the processor being programmed to:
create a color space transformation matrix,
transform the first color image to a second image using the transformation matrix, the second image having a plurality of long-wavelength (L), medium-wavelength (M), and short- wavelength (S) components,
determine white point adaptation levels for each of the plurality of long-wavelength (L), medium-wavelength (M), and short-wavelength (S) components,
adjust each of the plurality of long-wavelength (L), medium-wavelength (M), and short-wavelength (S) components by applying the associated white point adaptation levels to the plurality of long-wavelength (L), medium-wavelength (M), and short-wavelength (S) components,
stretch the adjusted long-wavelength (L) components and the short-wavelength (S) components with respect to the adjusted medium-wavelength (M) component,
apply the stretched L components and S components to the second image, and
transform the second image to an image in the first color space,
wherein each S component is stretched according to the equation S 1 =M+K(S 0 −M), and where K is a saturation adjustment factor, S 0 is an initial short-wavelength component, and S 1 is a stretched short-wavelength component.
6. The imaging device of claim 5 , wherein the first color image is an RGB image.
7. An imaging device comprising:
an array of pixels;
readout circuitry coupled to the array, the circuitry adapted to output a first color image of a first color space based on signals received from the array; and
an image processor adapted to input the first color image and perform saturation adjustment on the first color image, the processor being programmed to:
create a color space transformation matrix,
transform the first color image to a second image using the transformation matrix, the second image having a plurality of long-wavelength (L), medium-wavelength (M), and short-wavelength (S) components,
determine white point adaptation levels for each of the plurality of long-wavelength (L), medium-wavelength (M), and short-wavelength (S) components,
adjust each of the plurality of long-wavelength (L), medium-wavelength (M), and short-wavelength (S) components by applying the associated white point adaptation levels to the plurality of long-wavelength (L), medium-wavelength (M), and short-wavelength (S) components,
stretch the adjusted long-wavelength (L) components and the short-wavelength (S) components with respect to the adjusted medium-wavelength (M) component,
apply the stretched L components and S components to the second image, and
transform the second image to an image in the first color space,
wherein each L component is stretched according to the equation L 1 =M+K(L 0 −M), where K is a saturation adjustment factor, L 0 is an initial long-wavelength component, and L 1 is a stretched long-wavelength component.
8. The imaging device of claim 7 , wherein each S component is stretched according to the equation S 1 =M+K(S 0 −M), where S 0 is an initial short-wavelength component, and S 1 is a stretched short-wavelength component.
9. A storage medium containing a program, the program comprising instructions that when executed by a processor causes the processor to adjust a saturation of an LMS component image having a plurality of long-wavelength, medium-wavelength, and short-wavelength components by stretching the long-wavelength components and the short- wavelength components with respect to the medium-wavelength component,
wherein each long-wavelength component is stretched according to the equation L 1 =M+K(L 0 −M), where K is a saturation adjustment factor, M is a medium-wavelength component, L 0 is an initial long-wavelength component, and L 1 is a transformed long-wavelength component.
10. A storage medium containing a program, the program comprising instructions that when executed by a processor causes the processor to adjust a saturation of an LMS component image having a plurality of long-wavelength, medium-wavelength, and short-wavelength components by stretching the long-wavelength components and the short- wavelength components with respect to the medium-wavelength component,
wherein each short-wavelength component is stretched according to the equation S 1 =M+K(S 0 −M), where K is a saturation adjustment factor, M is a medium-wavelength component, S 0 is an initial short-wavelength component, and S 1 is a transformed short-wavelength component.