# US 2012/0250993 A1 · Image processing device, image processing method, and program

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2012/0250993 A1
- Document ID: US20120250993A1
- Status: Published application
- Kind: A1
- Application number: 13/425,155
- Filed: 2012-03-20
- Published: 2012-10-04
- Earliest claimed: 2011-03-29
- Assignees: Sony Corporation
- Inventors: Daisuke ISO; Tappei Kurita; Tomoo Mitsunaga
- Web page: https://patentgrep.com/patent/US20120250993A1

## CPC

- G06T 3/4015: Image demosaicing, e.g. colour filter arrays [CFA] or Bayer patterns
- G06T 5/70: Denoising; Smoothing
- G06T 5/73: Deblurring; Sharpening
- G06T 5/75: Unsharp masking
- G06T 5/77: Retouching; Inpainting; Scratch removal
- G06T 2200/12: involving antialiasing
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL

## Abstract

An image processing device includes an image estimation unit that estimates an image prior to the jaggedness occurrence from a jaggedness-occurring image and generates an estimated image prior to the jaggedness occurrence, and a weighting/adding unit that selects a jaggedness-occurring area as a processing target area in the estimated image prior to the jaggedness occurrence, detects a similar area that is a pixel area and similar to the processing target area, and then computes a weight according to the degree of similarity of each detected similar area to the processing target area, and detects a corresponding area in the jaggedness-occurring image to the processing target area and the similar area, and then computes a corrected pixel value of the processing target area of the jaggedness-occurring image through a weighting/adding process to which the weight of a pixel value of the detected corresponding area is applied.

## Claims

### Claim 1

An image processing device comprising:
an image estimation unit that estimates an image prior to the jaggedness occurrence from a jaggedness-occurring image and generates an estimated image prior to the jaggedness occurrence; and
a weighting/adding unit that selects a jaggedness-occurring area as a processing target area in the estimated image prior to the jaggedness occurrence, detects a similar area that is a pixel area and similar to the processing target area, and then computes a weight according to the degree of similarity of each detected similar area to the processing target area, and detects a corresponding area in the jaggedness-occurring image to the processing target area and the similar area, and then computes a corrected pixel value of the processing target area of the jaggedness-occurring image through a weighting/adding process to which the weight of a pixel value of the detected corresponding area is applied.

### Claim 2 (depends on 1)

The image processing device according to claim 1,
wherein the image estimation unit inputs a luminance image generated from the jaggedness-occurring image and generates an estimated luminance image prior to the jaggedness occurrence, and
wherein the weighting/adding unit computes the weight based on the estimated luminance prior to the jaggedness occurrence, and executes a weighting/adding process of pixel values to which the weight for the luminance image generated from the jaggedness-occurring image is applied.

### Claim 3 (depends on 1)

The image processing device according to claim 1, wherein the image estimation unit generates an estimated image prior to the jaggedness occurrence through a Bayes' estimation process to which preliminary learning data is applied.

### Claim 4 (depends on 1)

The image processing device according to claim 1, wherein the image estimation unit generates an estimated image prior to the jaggedness occurrence through a smoothing process for the jaggedness-occurring image.

### Claim 5 (depends on 1)

The image processing device according to claim 1, wherein the image estimation unit generates an estimated image prior to the jaggedness occurrence through a Bayes' estimation process to which statistical data of feature amounts expressing an edge is applied as preliminary learning data.

### Claim 6 (depends on 1)

The image processing device according to claim 1, wherein the weighting/adding unit executes a weighting/adding process to which an NL-Means (Non-Local-Means) method is applied.

### Claim 7 (depends on 1)

The image processing device according to claim 1, wherein the weighting/adding unit executes a weighting/adding process to which a Joint Bilateral Filter method is applied.

### Claim 8 (depends on 1)

The image processing device according to claim 1, further comprising:
a filtering process unit that generates corrected color difference information by executing a correlation process in application of the Joint Bilateral Filter (JBF), setting a jaggedness-reduced luminance image having a corrected pixel value computed by the weighting/adding unit to a reference image.

### Claim 9 (depends on 1)

The image processing device according to claim 1, further comprising:
a blending process unit that blends the jaggedness-reduced luminance image having a corrected pixel value computed by the weighting/adding unit and a jaggedness-occurring luminance image generated based on a jaggedness-occurring image prior to correction in accordance with a weight corresponding to an edge of an image.

### Claim 10

An image processing method for executing image processing in an image processing device, comprising:
estimating an image prior to the jaggedness occurrence from a jaggedness-occurring image and generating an estimated image prior to the jaggedness occurrence by an image estimation unit; and
selecting a jaggedness-occurring area as a processing target area in the estimated image prior to the jaggedness occurrence, detecting a similar area that is a pixel area and similar to the processing target area, and then computing a weight according to the degree of similarity of each detected similar area to the processing target area, and detecting a corresponding area in the jaggedness-occurring image to the processing target area and the similar area, and then computing a corrected pixel value of the processing target area of the jaggedness-occurring image through a weighting/adding process to which the weight of a pixel value of the detected corresponding area is applied, by a weighting/adding unit.

### Claim 11

A program that causes an image processing device to execute image processing, comprising:
estimating an image prior to the jaggedness occurrence from a jaggedness-occurring image and generating an estimated image prior to the jaggedness occurrence in an image estimation unit; and
selecting a jaggedness-occurring area as a processing target area in the estimated image prior to the jaggedness occurrence, detecting a similar area that is a pixel area and similar to the processing target area, and then computing a weight according to the degree of similarity of each detected similar area to the processing target area, and detecting a corresponding area in the jaggedness-occurring image to the processing target area and the similar area, and then computing a corrected pixel value of the processing target area of the jaggedness-occurring image through a weighting/adding process to which the weight of a pixel value of the detected corresponding area is applied, in a weighting/adding unit.
