Intelligent computational imaging system

1. A computer program product comprising logic instructions stored on a non-transitory computer readable medium, which when executed by a processor in an electronic device, configure the processor to implement intelligent computational imaging operations, comprising:

receiving a first set of input frames from an image capture device;

analyzing the first set of input frames to determine whether multi-image processing is to be implemented, and in response to a determination that multi-image processing is to be implemented:

determining a number of image frames to capture a dynamic range of a scene;

capturing at least the number of image frames;

aligning the number of images; and

merging the number of images into a merged image,

wherein determining a minimum number of image frames required to capture a scene comprises:

estimating a signal to noise ratio (SNR) of the first set of input frames;

estimating a dynamic range of the scene;

estimating a histogram of the spectral irradiance captured by the imaging device; and

incrementally increasing a number of image frames in a sample to maximize a sum of SNR of pixels captured.

2. The computer program product of claim 1 , further comprising logic instructions stored on a non-transitory computer readable medium, which when executed by a processor in an electronic device, configure the processor to process the merged image for display on a display device.

3. The computer program product of claim 1 , further comprising logic instructions stored on a non-transitory computer readable medium, which when executed by a processor in an electronic device, configure the processor to store the merged image in a memory module.

4. The computer program product of claim 1 , wherein analyzing the first set of input frames to determine whether multi-image processing is to be implemented comprises determining whether the entire dynamic range of the scene may be captured using a single exposure.

5. The computer program product of claim 1 , wherein estimating a dynamic range of the scene comprises:

determining a short exposure that represents bright regions of the scene by iteratively altering an exposure level until a threshold number of pixels fall within a pre-determined histogram range;

determining a long exposure that represents dark regions of the scene by iteratively altering the exposure level of the image until a threshold number of pixels fall within a predetermined histogram range.

6. The computer program product of claim 1 , wherein aligning the minimum number of images comprises:

estimating a rotation vector between images; and

warping the images to a common frame of reference.

7. The computer program product of claim 6 , wherein warping the images to a common frame of reference comprises applying a backwards interpolation method to the image frames in which red and blue channels are processed directly from the raw image frames, while green channels are interpolated on a grid rotated by 45 degrees.

8. The computer program product of claim 1 , wherein merging the minimum number of images into a merged image utilizes maximum likelihood combination.

9. The computer program product of claim 1 , further comprising logic instructions stored on a non-transitory computer readable medium, which when executed by a processor in an electronic device, configure the processor to bilinear demosaic and tonemap the merged image.

10. An electronic device, comprising:

an image capture device; and

logic to:

receive a first set of input frames from an image capture device;

analyze the first set of input frames to determine whether multi-image processing is to be implemented, and in response to a determination that multi-image processing is to be implemented:

determine a number of image frames required to capture a dynamic range of a scene;

capture at least the number of image frames;

align the number of images; and

merge the images into a merged image;

estimate a signal to noise ratio (SNR) of the first set of input frames;

estimate a dynamic range of the scene;

estimate a histogram of the spectral irradiance captured by the imaging device; and

incrementally increase a number of image frames in a sample to maximize a sum of SNR of pixels captured.

11. The electronic device of claim 10 , further comprising logic to determine whether the entire dynamic range of the scene may be captured using a single exposure.

12. The electronic device of claim 10 , further comprising logic to:

determine a short exposure that represents bright regions of the scene;

determine a long exposure that represents dark regions of the scene; and

iteratively increase an exposure level of the image until a threshold number of pixels fall within a predetermined histogram range.

13. The electronic device of claim 10 , further comprising logic to:

estimate a rotation vector between images; and

warp the images to a common frame of reference.

14. The electronic device of claim 10 , further comprising logic to merge the minimum number of images into a merged image utilizing a maximum likelihood combination.

15. The electronic device of claim 14 , further comprising logic to apply a backwards interpolation method to the raw image frames in which red and blue channels are processed directly from the raw image frames, while green channels are interpolated on a grid rotated by 45 degrees.

16. A method, comprising:

receiving a first set of input frames from an image capture device;

analyzing the first set of input frames to determine whether multi-image processing is to be implemented, and in response to a determination that multi-image processing is to be implemented:

determining a number of image frames to capture a dynamic range of a scene;

capturing at least the number of image frames;

aligning the number of images; and

merging the number of images into a merged image,

wherein determining a minimum number of image frames required to capture a scene comprises:

estimating a signal to noise ratio (SNR) of the first set of input frames;

estimating a dynamic range of the scene;

estimating a histogram of the spectral irradiance captured by the imaging device; and

incrementally increasing a number of image frames in a sample to maximize a sum of SNR of pixels captured.

17. The method of claim 16 , wherein estimating a dynamic range of the scene comprises:

determining a short exposure that represents bright regions of the scene by iteratively altering an exposure level until a threshold number of pixels fall within a pre-determined histogram range;

determining a long exposure that represents dark regions of the scene by iteratively altering the exposure level of the image until a threshold number of pixels fall within a predetermined histogram range.

18. An apparatus, comprising:

logic to:

receive a first set of input frames from an image capture device;

analyze the first set of input frames to determine whether multi-image processing is to be implemented, and in response to a determination that multi-image processing is to be implemented:

determine a number of image frames required to capture the full dynamic range of a scene;

capture at least the number of image frames;

align the number of images; and

merge the number of images into a merged image;

estimate a signal to noise ratio (SNR) of the first set of input frames;

estimate a dynamic range of the scene;

estimate a histogram of the spectral irradiance captured by the imaging device; and

incrementally increase a number of image frames in a sample to maximize a sum of SNR of pixels captured.

19. The apparatus of claim 18 , further comprising logic to process the merged image for display on a display device.

20. The apparatus of claim 18 , further comprising logic to store merged image in a memory module.

21. The apparatus of claim 18 , further comprising logic to determining whether the entire dynamic range of the scene may be captured with minimal visual noise using a single exposure.