Image processing apparatus and method of processing image for reducing noise of the image

• US 8,345,127 B2
• Filed: 06/20/2012
• Issued: 01/01/2013
• Est. Priority Date: 09/12/2008
• Status: Expired due to Fees

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Shohara, Makoto
• Primary Examiner(s): HERNANDEZ, NELSON D
• Application Number:
  US13/528,389
• Publication Number:
  US 20120257089A1
• Time in Patent Office: 195 Days
• Field of Search: 348/222.1, 348/272, 348/273, 348/241, 382260-266
• US Class Current: 348/241
• CPC Class Codes: G06T 2207/10024 Color image
• CPC Class Codes: G06T 2207/20012 Locally adaptive
• CPC Class Codes: G06T 2207/20192 Edge enhancement; Edge preservation
• CPC Class Codes: G06T 5/20 using local operators
• CPC Class Codes: G06T 5/70 Denoising; Smoothing
• CPC Class Codes: G06V 10/30 Noise filtering
• CPC Class Codes: H04N 2209/045 using mosaic colour filter
• CPC Class Codes: H04N 23/81 for suppressing or minimising disturbance in the image signal generation

Abstract

An image processing apparatus calculates a smoothed value obtained by smoothing signal levels of a plurality of pixels including a processing target pixel in a local area of an input image and a feature amount representing an edge degree of the processing target pixel using a pre-noise reduction image obtained by reducing an impulse noise of the input image. The image processing apparatus weighted-adds a signal level of the processing target pixel and the smoothed value at a ratio corresponding to the feature amount and outputs the weighted-addition result as a signal level after noise reduction processing.

Claims

1. An image processing apparatus comprising:

a reduced image generation unit configured to generate a reduced image of an input image;

a noise reduction unit configured to generate a first pre-noise reduction image obtained by reducing an impulse noise of the input image, and to generate a second pre-noise reduction image obtained by reducing the impulse noise of the reduced image;

a smoothing unit configured to calculate a first smoothed value obtained by smoothing signal levels of a plurality of pixels including a processing target pixel in a local area of the input image, and to calculate a second smoothed value obtained by smoothing signal levels of a plurality of pixels including the processing target pixel in a local area of the reduced image;

a feature amount calculation unit configured to calculate a first feature amount representing an edge degree of the processing target pixel using the first pre-noise reduction image, and to calculate a second feature amount representing an edge degree of the processing target pixel using the first pre-noise reduction image or the second pre-noise reduction image;

an addition unit configured to weighted-add a signal level of the processing target pixel of the input image and the first smoothed value at a ratio corresponding to the first feature amount in order to output the weighted addition result as a signal level after noise reduction processing of the input image, and to weighted-adds a signal level of the processing target pixel of the reduced image and the second smoothed value at a ratio corresponding to the second feature amount in order to output the weighted addition result as a signal level after noise reduction processing of the reduced image; and

a selection unit configured to select and output one of the signal level after noise reduction processing of the input image and the signal level after noise reduction processing of the reduced image according to a variance value of signal levels of the input image or a variance of signal levels of the first pre-noise reduction image of the input image,

wherein the addition unit performs the weighted addition such that as a pixel has a feature amount that represents a higher edge degree, a ratio of the smoothed value decreases.

2. The image processing apparatus according to claim 1, wherein the selection unit selects and outputs the signal level after noise reduction processing of the input image when the variance value of signal levels of the input image is a threshold value or more, and selects and outputs the signal level after noise reduction processing of the reduced image when the variance value of signal levels of the input image is smaller than the threshold value or selects and outputs the signal level after noise reduction processing of the input image when the variance value of the first pre-noise reduction image of the input image is a threshold value or more, and selects and outputs the signal level after noise reduction processing of the reduced image when the variance value of the first pre-noise reduction image of the input image is smaller than the threshold value.

3. The image processing apparatus according to claim 2, wherein the threshold value is a variance value of noises included in the signal level of the reduced image.

4. The image processing apparatus according to claim 3, further comprising a mixing processing unit configured to weighted-add the signal level output from the selection unit and a signal level of the input image and to output a signal level of the weighted-addition result.

5. The image processing apparatus according to claim 4, wherein the mixing processing unit performs a weighted addition such that as a sensitivity set in an imaging apparatus when the input image is captured is higher, a ratio of the signal level output from the selection unit is higher.

6. The image processing apparatus according to claim 1, wherein the first feature amount calculation unit calculates the first feature amount using a variance value of signal levels of a plurality of pixels including a processing target pixel in a local area of the first pre-noise reduction image and a variance value of noises of the first pre-noise reduction image, and

wherein as a ratio of the variance value of the signal levels to the variance value of noises increases, the edge degree represented by the feature amount increases.

7. An image processing method comprising:

generating a reduced image of an input image;

generating a first pre-noise reduction image obtained by reducing an impulse noise of the input image;

generating a second pre-noise reduction image obtained by reducing an impulse noise of the reduced image;

calculating a first smoothed value obtained by smoothing signal levels of a plurality of pixels including a processing target pixel in a local area of the input image;

calculating a second smoothed value obtained by smoothing signal levels of a plurality of pixels including a processing target pixel in a local area of the reduced image;

calculating a first feature amount representing an edge degree of the processing target pixel using the first pre-noise reduction image;

calculating a second feature amount representing an edge degree of the processing target pixel using the first pre-noise reduction image or the second pre-noise reduction image; and

weighted-adding a signal level of the processing target pixel of the input image and the first smoothed value at a ratio corresponding to the first feature amount and outputting the weighted addition result as a signal level after noise reduction processing of the input image,

weighted-adding a signal level of the processing target pixel of the reduced image and the second smoothed value at a ratio corresponding to the second feature amount and outputting the weighted addition result as a signal level after noise reduction processing of the reduced image,

selecting and outputting one of the signal level after noise reduction processing of the input image and the signal level after noise reduction processing of the reduced image according to a variance value of signal levels of the input image or a variance of signal levels of the first pre-noise reduction image of the input image,

wherein in weighted-adding the signal level, as a pixel has a feature amount that represents a higher edge degree, a ratio of the smoothed value decreases.

8. A non-transitory computer-readable medium storing a program that, when executed by a computer, causes the computer to perform a method, the method comprising the steps of:

generating a reduced image of an input image;

generating a first pre-noise reduction image obtained by reducing an impulse noise of the input image;

generating a second pre-noise reduction image obtained by reducing an impulse noise of the reduced image;

calculating a first smoothed value obtained by smoothing signal levels of a plurality of pixels including a processing target pixel in a local area of the input image;

calculating a second smoothed value obtained by smoothing signal levels of a plurality of pixels including a processing target pixel in a local area of the reduced image;

calculating a first feature amount representing an edge degree of the processing target pixel using the first pre-noise reduction image;

calculating a second feature amount representing an edge degree of the processing target pixel using the first pre-noise reduction image or the second pre-noise reduction image; and

weighted-adding a signal level of the processing target pixel of the input image and the first smoothed value at a ratio corresponding to the first feature amount and outputting the weighted addition result as a signal level after noise reduction processing of the input image,

weighted-adding a signal level of the processing target pixel of the reduced image and the second smoothed value at a ratio corresponding to the second feature amount and outputting the weighted addition result as a signal level after noise reduction processing of the reduced image,

selecting and outputting one of the signal level after noise reduction processing of the input image and the signal level after noise reduction processing of the reduced image according to a variance value of signal levels of the input image or a variance of signal levels of the first pre-noise reduction image of the input image,

wherein in weighted-adding the signal level, as a pixel has a feature amount that represents a higher edge degree, a ratio of the smoothed value decreases.