Estimating noise for a digital image utilizing updated statistics
View Patent ↗A method of estimating noise in a digital image receiving a digital image; providing a first statistical table related to noise provided by an image source; and using the pixels of the digital image to calculate a second statistical table related to the noise provided by the image source in capturing the image. The method further includes using the first statistical table and the second statistical table to provide an updated third statistical table related to the noise provided by the image source in capturing the image; and calculating a noise characteristic table from the third statistical table for use in enhancing the digital image.
1. A method of estimating noise in a digital image comprising:
accumulating statistical noise data in a plurality of default statistical tables, wherein each default statistical table corresponds to a unique source identification tag associated with a particular film type, and wherein the statistical noise data within a given default statistical table is related to a particular film type;
utilizing image pixel data from a digital image to calculate a statistical noise data table corresponding to the digital image;
utilizing a source identification tag corresponding to the digital image to select a default statistical table from the plurality of default statistical tables; and
utilizing the selected default statistical table in conjunction with the statistical noise data table corresponding to the digital image to generate a resultant calculated noise characteristic table.
2. The method of claim 1 , wherein the default statistical tables, the statistical noise data table and the resultant calculated noise characteristic table each includes a series of standard deviation values for different ranges of scanner intensities.
3. The method claimed in claim 2 , wherein the resultant calculated noise characteristic table is used in processing the digital image to generate an enhanced digital image.
4. The method claimed in claim 3 , wherein a spatial filter is used to calculate the enhanced digital image.
5. The method claimed in claim 3 , further including the step of using the resultant calculated noise characteristic table and a noise reduction filter to calculate the enhanced digital image.
6. The method claimed in claim 3 , further including the step of using the resultant calculated noise characteristic table and a spatial sharpening filter to calculate the enhanced digital image.
7. The method claimed in claim 3 , further including the step of using the resultant calculated noise characteristic table, a noise reduction filter and a spatial sharpening filter to calculate the enhanced digital image.
8. The method of claim 1 , wherein default statistical tables, the statistical noise data table and the resultant calculated noise characteristic table each includes at least one histogram.
9. The method of claim 1 , wherein default statistical tables, the statistical noise data table and the resultant calculated noise characteristic table each includes a series of histograms for different ranges of scanner intensities.
10. A method as claimed in claim 1 , wherein the digital image is generated by a scanning device capable of scanning a film and generating image pixel data corresponding to the digital image and a source identification tag corresponding to the digital image.