Transforming a digital image from a low dynamic range (LDR) image to a high dynamic range (HDR) image

1. A method for transforming a digital image from a Low Dynamic Range (LDR) image obtained with a given camera to a High Dynamic Range (HDR) image, comprising:

obtaining an exposure-pixel response curve for the given camera;

converting the LDR image to hue, saturation, and brightness (HSB) color space arrays including a hue array, a saturation array, and a brightness array; and

generating a radiance array by inverse mapping each pixel in the brightness array using an inverse of the exposure-pixel response curve.

2. The method of claim 1 , further comprising:

for each pixel in the radiance array determining a local luminance average;

generating an adjusted radiance array by adjusting each pixel of the radiance array using the local luminance average for the pixel; and

converting the LDR image, using the hue array, the saturation array, and the adjusted radiance array, into the HDR image.

3. The method of claim 2 , wherein

for each pixel (u,v), the adjusted radiance radiance′[u,v] is computed according to:

Radiance′[ u,v ]=Radiance[ u,v ]*Radiance[ u,v ]/Local_Luminance_Average[ u,v ], where

Radiance[u,v] designates a radiance value for the pixel (u,v) and

Local_Luminance_Average[u,v] designates the local luminance average at the pixel (u,v).

4. The method of claim 2 , wherein

the local luminance average is determined from i convolution kernels defined as:

kernel i [ ]=GaussianKernel( r i )[ ], where

r i is a radius of a GaussianKernel and,

i designates a local contrast scale index varying from 0 to 8.

5. The method of claim 4 , wherein

r i is 1/(2*Sqrt(2))*1.6 i .

6. The method of claim 1 , wherein

the LDR image is converted according to:

Convert(Hue[ ],Saturation[ ],Radiance[ ]/max(Radiance[ ]))*max(Radiance), where

Hue[ ] designates the hue array,

Saturation[ ] the saturation array, and

Radiance[ ] the radiance array.

7. The method of claim 1 , wherein

the exposure-pixel response curve is obtained from a camera data sheet of the given camera.

8. The method of claim 1 , wherein

the exposure-pixel response curve is obtained using a sequence of differently exposed images from the given camera and for a same screen.

9. A computer hardware system configured to transforms a digital image from a Low Dynamic Range (LDR) image obtained with a given camera to a High Dynamic Range (HDR) image, comprising:

a processor, wherein the processor is configured to perform

obtaining an exposure-pixel response curve for the given camera;

converting the LDR image to hue, saturation, and brightness (HSB) color space arrays including a hue array, a saturation array, and a brightness array; and

generating a radiance array by inverse mapping each pixel in the brightness array using an inverse of the exposure-pixel response curve.

10. The computer hardware system of claim 9 , wherein the processor is further configured to perform

for each pixel in the radiance array determining a local luminance average;

generating an adjusted radiance array by adjusting each pixel of the radiance array using the local luminance average for the pixel; and

converting the LDR image, using the hue array, the saturation array, and the adjusted radiance array, into the HDR image.

11. The computer hardware system of claim 10 , wherein

for each pixel (u,v), the adjusted radiance radiance′[u,v] is computed according to:

Radiance′[ u,v ]=Radiance[ u,v ]*Radiance[ u,v ]/Local_Luminance_Average[ u,v ], where

Radiance[u,v] designates a radiance value for the pixel (u,v) and

Local_Luminance_Average[u,v] designates the local luminance average at the pixel (u,v).

12. The computer hardware system of claim 10 , wherein

the local luminance average is determined from i convolution kernels defined as:

kernel i [ ]=GaussianKernel( r i )[ ],where

r i is a radius of a GaussianKernel and,

i designates a local contrast scale index varying from 0 to 8.

13. The computer hardware system of claim 12 , wherein

r i is 1/(2*Sqrt(2))*1.6 i .

14. The computer hardware system of claim 9 , wherein

the LDR image is converted according to:

Convert(Hue[ ],Saturation[ ],Radiance[ ]/max(Radiance[ ]))*max(Radiance), where

Hue[ ] designates the hue array,

Saturation[ ] the saturation array, and

Radiance[ ] the radiance array.

15. A computer program product comprising a computer usable storage medium having stored therein computer usable program code for transforming a digital image from a Low Dynamic Range (LDR) image obtained with a given camera to a High Dynamic Range (HDR) image, the computer usable program code, which when executed by a computer hardware system, causes the computer hardware system to perform:

obtaining an exposure-pixel response curve for the given camera;

converting the LDR image to hue, saturation, and brightness (HSB) color space arrays including a hue array, a saturation array, and a brightness array; and

generating a radiance array by inverse mapping each pixel in the brightness array using an inverse of the exposure-pixel response curve, wherein

the computer usable storage medium is not a transitory, propagating signal per se.

16. The computer program product of claim 15 , further comprising:

for each pixel in the radiance array determining a local luminance average;

generating an adjusted radiance array by adjusting each pixel of the radiance array using the local luminance average for the pixel; and

converting the LDR image, using the hue array, the saturation array, and the adjusted radiance array, into the HDR image.

17. The computer program product of claim 16 , wherein

for each pixel (u,v), the adjusted radiance radiance′[u,v] is computed according to:

Radiance′[ u,v ]=Radiance[ u,v ]*Radiance[ u,v ]/Local_Luminance_Average[ u,v ], where

Radiance[u,v] designates a radiance value for the pixel (u,v) and

Local_Luminance_Average[u,v] designates the local luminance average at the pixel (u,v).

18. The computer program product of claim 16 , wherein

the local luminance average is determined from i convolution kernels defined as:

kernel i [ ]=GaussianKernel( r i )[ ],where

r i is a radius of a GaussianKernel and,

i designates a local contrast scale index varying from 0 to 8.

19. The computer program product of claim 18 , wherein

r i is 1/(2*Sqrt(2))*1.6 i .

20. The computer program product of claim 15 , wherein

the LDR image is converted according to:

Convert(Hue[ ],Saturation[ ],Radiance[ ]/max(Radiance[ ]))*max(Radiance), where

Hue[ ] designates the hue array,

Saturation[ ] the saturation array, and

Radiance[ ] the radiance array.