Back to results
US 8,265,412 B2Granted
Image processing method
AssigneeNikon Corporation
InventorKenichi Ishiga
Life of this patent
Estimating the term…
Abstract
An image processing method for removing a noise component contained in an original image includes: smoothing a noise component contained in an original image on a temporary basis by using a noise fluctuation index value; extracting a temporary noise component free of a non-noise component based upon a differential signal representing a difference between the original image and a smoothed image; extracting an actual noise component by further excluding a component less likely to be a noise component from the extracted temporary noise component by comparing the extracted temporary noise component with the noise fluctuation index value again; and removing noise from the original image based upon the extracted actual noise component having been extracted.
Claims24
An image processing method for removing a noise component contained in an original image, comprising:
- smoothing a noise component contained in an original image on a temporary basis by using a noise fluctuation index value;
- extracting a temporary noise component free of a non-noise component based upon a differential signal representing a difference between the original image and a smoothed image;
- extracting an actual noise component by further excluding a component less likely to be a noise component from the extracted temporary noise component by comparing the extracted temporary noise component with the noise fluctuation index value again;
- and
- removing noise from the original image based upon the extracted actual noise component having been extracted.
2 more elements
An image processing method for removing a noise component contained in an original image, comprising:
- adaptively smoothing a noise component contained in an original image by distinguishing the noise component from an edge component present within a smoothing target area through comparison of a noise fluctuation index value and a spatial fluctuation width over which original image signals fluctuate;
- extracting a temporary noise component based upon a differential signal representing a difference between the original image and an image resulting from adaptive smoothing;
- statistically re-examining a distribution of the temporary noise component having been extracted by comparing the extracted temporary noise component with the noise fluctuation index value again;
- extracting an actual noise component by further excluding a component less likely to be a noise component from the extracted temporary noise component based upon examination results;
- and
- removing noise from the original image based upon the actual noise component having been extracted.
3 more elements
An image processing method for removing a noise component contained in an original image, comprising:
- smoothing an original image by using a specific noise model having a distribution function;
- extracting a temporary noise component based upon a differential signal representing a difference between the original image and the image having been smoothed;
- estimating a quantity of false noise component present in the temporary noise component having been extracted by comparing a value indicated for the temporary noise component with the distribution function of the specific noise model;
- extracting an actual noise component by excluding the false noise component in the estimated quantity from the temporary noise component;
- and
- removing noise from the original image based upon the actual noise component having been extracted.
3 more elements
An image processing method for removing a noise component contained in an original image, comprising:
- generating a plurality of band-limited images assuming gradually lowered resolution levels by filtering the original image;
- extracting a noise component free of a non-noise component in each band-limited image by executing noise removal processing by comparing the band-limited image with a noise fluctuation index value determined in correspondence to the band-limited image;
- synthesizing noise component signals having been extracted from the band-limited images to generate a noise component assuming a resolution level matching the resolution level of the original image;
- estimating a quantity of a false noise component present in the synthesized noise component by comparing a level of the synthesized noise component with the noise fluctuation index value determined in correspondence to the original image again;
- extracting an actual noise component by excluding the false noise component from the synthesized noise component based upon estimation results;
- and
- removing noise from the original image based upon the actual noise component having been extracted.
4 more elements
An image processing method for removing noise component contained in an original image, comprising:
- generating a plurality of band-limited images assuming gradually lowered resolution levels by filtering the original image;
- extracting a noise component free of a non-noise component in each band-limited image by executing noise removal processing by comparing the band-limited image with a noise fluctuation index value determined in correspondence to the band-limited image;
- synthesizing noise component signals having been extracted from the band-limited images to generate a noise component assuming a resolution level matching the resolution level of the original image;
- estimating a quantity of a false noise component present in the synthesized noise component by comparing a level of the synthesized noise component with the noise fluctuation index value determined in correspondence to the original image again;
- extracting an actual noise component by excluding the false noise component from the synthesized noise component based upon estimation results;
- removing noise from the original image based upon the actual noise component having been extracted;
- estimating the quantity of false noise component present in the noise component by comparing the level of the noise component free of a non-noise component, having been extracted in correspondence to each of the band-limited images, with the noise fluctuation index value determined in correspondence to the band-limited image again;
- extracting the actual noise component in each band-limited image by excluding the false noise component from the noise component based upon estimation results;
- and
- synthesizing individual actual noise component signals having been extracted so that verification is executed to exclude the false noise component twice by excluding the false noise component in each band-limited image prior to noise component synthesis and also by excluding the false noise component at a resolution level matching the resolution of the original image following the noise component synthesis.
7 more elements
Dependencies are inferred from the claim text.