Back to results
US 12,217,392 B1Granted
Techniques for preserving original luma with noisy clamped sources
AssigneeAmazon Technologies, Inc.
InventorCharles Benjamin Waggoner
Life of this patent
Estimating the term…
Abstract
Techniques are disclosed for correcting noisy clamped image sources while preserving original luma, chroma, or other channel value. A computing device can receive an image that has a plurality of pixels associated with respective channel values. The computing device can then determine a region of the image for correction and generate a histogram of the respective channel values in the region. Based on the histogram, the computing device can generate a corrected histogram of corrected channel values. The computing device can then denoise the corrected channel values.
Claims20
A method performed by a computing device, the method comprising:
- receiving, by the computing device, an image comprising a plurality of pixels, each pixel of the plurality of pixels associated with a respective luma value of a plurality of respective luma values;
- determining, by the computing device, a region within the image, the region comprising a subset of the plurality of pixels of the image;
- generating, by the computing device, a histogram of the plurality of respective luma values of the subset of the plurality of pixels in the region, the histogram corresponding to a distribution of the plurality of respective luma values in the region, and the distribution characterized by normally-distributed noise in the luma values;
- identifying, by the computing device, a luma level bin of the histogram containing a number of the subset of the plurality of pixels exceeding a threshold number, the luma level bin associated with a minimum luma value of the distribution;
- generating, using the histogram and based at least in part on the luma level bin, a corrected histogram comprising corrected luma values for the subset of the pixels, at least one of the corrected luma values being less than the minimum luma value;
- denoising, by the computing device, the corrected luma values by reducing a variance in the corrected luma values to produce denoised luma values for the subset of the plurality of pixels in the region;
- and
- generating, by the computing device, a corrected image region of a corrected image comprising a plurality of corrected pixels having the denoised luma values.
5 more elements
A computing device, comprising:
- one or more processors;
- and
- one or more memories storing computer-executable instructions that, when executed by the one or more processors cause the computing device to:
- receive an image comprising a plurality of pixels, each pixel of the plurality of pixels associated with a respective channel value of a plurality of respective channel values;
- determine a region within the image, the region comprising a subset of the plurality of pixels of the image;
- generate a histogram of the plurality of respective channel values of the subset of the plurality of pixels in the region, the histogram corresponding to a distribution of the plurality of respective channel values in the region;
- generate, using the histogram and based at least in part on a channel level bin of the histogram containing a number of the subset of the plurality of pixels exceeding a threshold number, a corrected histogram comprising corrected channel values for the subset of the plurality of pixels;
- and
- denoise the corrected channel values to produce denoised channel values for the subset of the plurality of pixels in the region.
6 more elements
A non-transitory computer-readable storage medium storing computer-executable instructions that, when executed by one or more processors of a computer system, cause the computer system to at least:
- receive, by the computer system, an image comprising a plurality of pixels, each pixel of the plurality of pixels associated with a respective channel value of a plurality of respective channel values;
- determine, by the computer system, a region within the image, the region comprising a subset of the plurality of pixels of the image;
- generate, by the computer system, a histogram of the respective channel values of the subset of the plurality of pixels in the region, the histogram corresponding to a distribution of the respective channel values in the region, the distribution characterized by normally-distributed noise in the respective channel values;
- identify, by the computer system using the histogram, a channel level bin containing a number of the subset of the plurality of pixels in excess of a threshold number;
- generate, by the computer system using the histogram and based at least in part on the channel level bin, a corrected histogram comprising corrected channel values for the subset of the plurality of pixels;
- denoising, by the computer system, the corrected channel values to produce denoised channel values for the subset of the plurality of pixels in the region;
- and
- generating, by the computer system, a corrected image region of a corrected image comprising a plurality of corrected pixels having the denoised channel values.
5 more elements
Dependencies are inferred from the claim text.