# US 2012/0269430 A1 · System and method for combining color information with spatial information in multispectral images

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

- Number: US 2012/0269430 A1
- Document ID: US20120269430A1
- Status: Published application
- Kind: A1
- Application number: 13/452,011
- Filed: 2012-04-20
- Published: 2012-10-25
- Earliest claimed: 2011-04-22
- Assignees: Exelis, Inc.
- Inventors: Michael Paul Deskevich; Christopher Padwick
- Web page: https://patentgrep.com/patent/US20120269430A1

## CPC

- G06T 5/73: Deblurring; Sharpening
- G06F 18/251: {of input or preprocessed data}
- G06V 10/803: {of input or preprocessed data}
- G06V 20/13: Satellite images
- G06T 2207/10024: Color image
- G06T 2207/10036: Multispectral image; Hyperspectral image
- G06T 2207/10041: Panchromatic image
- G06T 2207/20021: Dividing image into blocks, subimages or windows
- G06T 2207/20221: Image fusion; Image merging
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING

## Abstract

An image processing system combines higher-resolution panchromatic images and lower resolution multispectral images using a hyperspherical color space pan-sharpening technique. By converting the multispectral images into a hyperspherical color space, the intensities of the multispectral images can be intensity matched to the intensities of the panchromatic image and then retransformed back to the original color space. The intensity matching can utilize a number of techniques, including, but not limited to, direct substitution of the intensities of the panchromatic image for the intensities of the multispectral images, modification of the intensities of the multispectral images based on predefined statistical models and modification of the intensities of the multispectral images based on dynamically generated statistical models and a selected sharpening parameter β.

## Claims

### Claim 1

An image processing system for combining at least a portion of a panchromatic image block at a first resolution with at least corresponding cogeolocated portions of n spectral band image blocks of a second resolution lower than the first resolution, where n is an integer that is at least as large as two, the system comprising:
circuitry for upsampling the corresponding portions of the n spectral band image blocks at the second resolution to form n upsampled spectral band image blocks;
circuitry for transforming the n upsampled spectral band image blocks to a hyperspherical color space to form hyperspherical color space image blocks including, for each pixel in the hyperspherical color space image block, a hyperspherical intensity and n-1 hyperspherical angles;
circuitry for intensity matching the intensities of the hyperspherical color space image blocks with intensities of corresponding pixels in the panchromatic image block to form intensity adjusted hyperspherical color space image blocks; and
circuitry for reverse transforming the intensity adjusted hyperspherical color space image blocks into n reverse transformed spectral band image blocks.

### Claim 2 (depends on 1)

The image processing circuitry as claimed in claim 1, wherein the circuitry for upsampling comprises circuitry for performing a zero order hold on the n spectral band image blocks of a second resolution to generate pixels at the first resolution.

### Claim 3 (depends on 1)

The image processing circuitry as claimed in claim 1, wherein the circuitry for upsampling comprises circuitry for performing linear interpolation on the n spectral band image blocks of a second resolution to generate pixels at the first resolution.

### Claim 4 (depends on 1)

The image processing circuitry as claimed in claim 1, wherein the circuitry for upsampling comprises circuitry for performing cubic interpolation on the n spectral band image blocks of a second resolution to generate pixels at the first resolution.

### Claim 5 (depends on 1)

The image processing circuitry as claimed in claim 1, wherein the circuitry for upsampling comprises circuitry for performing a non-linear interpolation on the n spectral band image blocks of a second resolution to generate pixels at the first resolution.

### Claim 6 (depends on 1)

The image processing circuitry as claimed in claim 1, wherein the circuitry for intensity matching the intensities of the hyperspherical color space image blocks with the intensities of corresponding pixels in the panchromatic image block comprises circuitry for using the intensities of corresponding pixels in the panchromatic image block as the intensity adjusted hyperspherical color space image blocks.

### Claim 7 (depends on 1)

The image processing circuitry as claimed in claim 1, wherein the circuitry for intensity matching the intensities of the hyperspherical color space image blocks with the intensities of corresponding pixels in the panchromatic image block comprises circuitry for filtering at least the intensities of corresponding pixels in the panchromatic image block to form the intensity adjusted hyperspherical color space image blocks.

### Claim 8 (depends on 1)

The image processing circuitry as claimed in claim 1, wherein the circuitry for intensity matching the intensities of the hyperspherical color space image blocks with the intensities of corresponding pixels in the panchromatic image block comprises circuitry for smoothing at least the intensities of the corresponding pixels in the panchromatic image block to form the intensity adjusted hyperspherical color space image blocks.

### Claim 9 (depends on 1)

The image processing circuitry as claimed in claim 1, wherein the circuitry for intensity matching the intensities of the hyperspherical color space image blocks with the intensities of corresponding pixels in the panchromatic image block comprises circuitry for forming the intensity adjusted hyperspherical color space image blocks by filtering at least the intensities of corresponding pixels in the panchromatic image block using statistical characteristics of at least one of (a) the intensities of corresponding pixels in the panchromatic image block and (b) the intensities of the hyperspherical color space image blocks.

### Claim 10 (depends on 9)

The image processing circuitry as claimed in claim 9, wherein the statistical characteristics comprises at least one of (a) a mean of the square of the intensities of corresponding pixels in the panchromatic image block, (b) a standard deviation of the square of the intensities of corresponding pixels in the panchromatic image block, (c) a mean of the square of the intensities of the hyperspherical color space image blocks and (d) a standard deviation of the square of the intensities of the hyperspherical color space image blocks.

### Claim 11 (depends on 1)

The image processing circuitry as claimed in claim 1, wherein the circuitry for intensity matching the intensities (I) of the hyperspherical color space image blocks with the intensities (P) of corresponding pixels in the panchromatic image block comprises circuitry for forming the intensity adjusted hyperspherical color space image blocks (Iadj) by calculating:
P
′2

(
x
,
y
)
=
σ
i
σ
p

(
P

(
x
,
y
)
2
-
μ
p
+
σ
p
)
+
μ
i
-
σ
i
,
and
I
adj

(
x
,
y
)
=
P
′2

(
x
,
y
)
,
where P2(x,y)2 is the square of the intensity of the corresponding pixel at location (x,y) in the panchromatic image block, μp and μp are the mean and standard deviation of the square of the intensities of the corresponding pixels in the panchromatic image block (P2), respectively, and μi and σi are the mean and standard deviation of I2.

### Claim 12

A processor-based method for combining at least a portion of a panchromatic image block at a first resolution with at least corresponding cogeolocated portions of n spectral band image blocks of a second resolution lower than the first resolution, where n is an integer that is at least as large as two, the processor performing instructions stored in a computer memory to cause the processor to perform the method comprising:
upsampling the corresponding portions of the n spectral band image blocks at the second resolution to form n upsampled spectral band image blocks;
transforming the n upsampled spectral band image blocks to a hyperspherical color space to form hyperspherical color space image blocks including, for each pixel in the hyperspherical color space image block, a hyperspherical intensity and n-1 hyperspherical angles;
intensity matching the intensities of the hyperspherical color space image blocks with intensities of corresponding pixels in the panchromatic image block to form intensity adjusted hyperspherical color space image blocks; and
reverse transforming the intensity adjusted hyperspherical color space image blocks into n reverse transformed spectral band image blocks.

### Claim 13 (depends on 12)

The processor-based method as claimed in claim 12, wherein the intensity matching the intensities of the hyperspherical color space image blocks with the intensities of corresponding pixels in the panchromatic image block comprises using the intensities of corresponding pixels in the panchromatic image block as the intensity adjusted hyperspherical color space image blocks.

### Claim 14 (depends on 12)

The processor-based method as claimed in claim 12, wherein the intensity matching the intensities of the hyperspherical color space image blocks with the intensities of corresponding pixels in the panchromatic image block comprises forming the intensity adjusted hyperspherical color space image blocks by filtering at least the intensities of corresponding pixels in the panchromatic image block using statistical characteristics of at least one of (a) the intensities of corresponding pixels in the panchromatic image block and (b) the intensities of the hyperspherical color space image blocks.

### Claim 15 (depends on 14)

The processor-based method as claimed in claim 14, wherein the statistical characteristics comprises at least one of (a) a mean of the square of the intensities of corresponding pixels in the panchromatic image block, (b) a standard deviation of the square of the intensities of corresponding pixels in the panchromatic image block, (c) a mean of the square of the intensities of the hyperspherical color space image blocks and (d) a standard deviation of the square of the intensities of the hyperspherical color space image blocks.

### Claim 16 (depends on 12)

The processor-based method as claimed in claim 12, wherein the intensity matching the intensities (I) of the hyperspherical color space image blocks with the intensities (P) of corresponding pixels in the panchromatic image block comprises forming the intensity adjusted hyperspherical color space image blocks (Iadj) by calculating:
P
′2

(
x
,
y
)
=
σ
i
σ
p

(
P

(
x
,
y
)
2
-
μ
p
+
σ
p
)
+
μ
i
-
σ
i
,
and
I
adj

(
x
,
y
)
=
P
′2

(
x
,
y
)
,
where P2(x,y)2 is the square of the intensity of the corresponding pixel at location (x,y) in the panchromatic image block, μp and σp are the mean and standard deviation of the square of the intensities of the corresponding pixels in the panchromatic image block (P2), respectively, and μi and σi are the mean and standard deviation of I2.
