# US 2012/0075506 A1 · Noise reduction for machine vision systems

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

- Number: US 2012/0075506 A1
- Document ID: US20120075506A1
- Status: Published application
- Kind: A1
- Application number: 13/100,060
- Filed: 2011-05-03
- Published: 2012-03-29
- Earliest claimed: 2010-09-27
- Assignees: Sharp Laboratories of America, Inc.
- Inventors: Petrus J. L. van Beek
- Web page: https://patentgrep.com/patent/US20120075506A1

## CPC

- G06T 5/20: using local operators
- G06T 5/70: Denoising; Smoothing
- H04N 23/56: provided with illuminating means
- H04N 25/671: for non-uniformity detection or correction
- G06T 2207/10144: Varying exposure
- G06T 2207/10152: Varying illumination
- G06T 2207/20032: Median filtering
- G06T 2207/20182: Noise reduction or smoothing in the temporal domain; Spatio-temporal filtering
- G06T 2207/20224: Image subtraction
- H04N 25/618: for random or high-frequency noise
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION

## Abstract

A method for reducing noise in an image from an image capture device includes filter the image using both an offset fixed pattern noise filter and a gain fixed pattern noise filter. Thereafter, the image is filtered using a remnant fixed pattern noise filter to reduce remnant fixed pattern noise.

## Claims

### Claim 1

A method for reducing noise in an image from an image capture device comprising:
(a) receiving an image from an image capture device;
(b) processing said image in a manner to reduce offset fixed pattern noise in said image;
(c) processing said image in a manner to reduce gain fixed pattern noise in said image;
(d) filtering said filtered image as a result of step (b) and step (c) using a remnant fixed pattern noise filter to reduce remnant fixed pattern noise, wherein said remnant fixed pattern noise filter is characterized by including an impulse.

### Claim 2 (depends on 1)

The method of claim 1 wherein said remnant fixed pattern noise filter is further characterized by including a rank of pixels within a window.

### Claim 3 (depends on 2)

The method of claim 2 wherein said window is a 3×3 window.

### Claim 4 (depends on 2)

The method of claim 2 wherein said rank of pixels is in a sequential rank order.

### Claim 5 (depends on 4)

The method of claim 4 wherein a center pixel of said window is selective modified based upon said remnant fixed pattern noise filter.

### Claim 6 (depends on 5)

The method of claim 5 wherein said selective modification includes a condition that said center pixel has a minimum or a maximum of said pixels of said window.

### Claim 7 (depends on 6)

The method of claim 6 wherein said selective modification includes a condition based upon pixel value differences.

### Claim 8 (depends on 7)

The method of claim 7 wherein said pixel value differences is based upon said center pixel being sufficiently different from all other pixel values in said window.

### Claim 9 (depends on 8)

The method of claim 8 wherein said pixel value differences is further based upon said all other pixel values being sufficiently similar to each other.

### Claim 10 (depends on 9)

The method of claim 9 wherein said remnant fixed pattern noise filter is characterized by:
f
^
c
←
{
r
2


if
(
f
5
=
r
1
)
⋀
(
r
2
-
r
1
>
T
RD
)
⋀
(
(
r
2
-
r
8
)
2
+
(
r
2
-
r
9
)
2
<
T
RV
)
r
8


if
(
f
5
=
r
9
)
⋀
(
r
9
-
r
8
>
T
RD
)
⋀
(
(
r
8
-
r
1
)
2
+
(
r
8
-
r
2
)
2
<
T
RV
)
f
5
otherwise

### Claim 11

A method for reducing noise in an image from an image capture device comprising:
(a) receiving an image from an image capture device;
(b) processing said image in a manner to reduce offset fixed pattern noise in said image;
(c) processing said image in a manner to reduce gain fixed pattern noise in said image;
(d) filtering said filtered image as a result of step (b) and step (c) using a remnant fixed pattern noise filter to reduce remnant fixed pattern noise;
(e) filtering the result of step (d) using a random noise filter that selective modifies a center pixel, wherein said random noise filter is characterized by including a first window centered on said center pixel and a second smaller window centered on said center pixel, wherein said random noise filter uses a pair of filters, a first filter based upon said pixels in said second window and a second filter based upon pixels in said first window and not in said first window.

### Claim 12 (depends on 11)

The method of claim 11 wherein said random noise filter is further characterized by said second filter being a mean based filter.

### Claim 13 (depends on 12)

The method of claim 12 wherein said random noise filter is further characterized by said first filter being a variance based filter.

### Claim 14 (depends on 13)

The method of claim 13 wherein said variance based filter is based upon differences between pixel value in the second window and said mean based filter.
