# US 8,150,150 B2 · Method and system of extracting a perceptual feature set

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

- Number: US 8,150,150 B2
- Document ID: US8150150B2
- Status: Granted
- Kind: B2
- Application number: 12/396,380
- Filed: 2009-03-02
- Published: 2010-09-02
- Granted: 2012-04-03
- Earliest claimed: 2009-03-02
- Assignees: Himax Technologies Limited; NCKU Research and Development Foundation
- Inventors: Gwo Giun Lee; Ming-Jiun Wang; Ling-Hsiu Huang
- Web page: https://patentgrep.com/patent/US8150150B2

## CPC

- G06V 10/56: relating to colour
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING

## Abstract

A method and system of extracting a perceptual feature set for image/video segmentation are disclosed. An input image is converted to obtain a hue component and a saturation component, where the hue component is quantized into a number of quantum values. After weighting the quantized hue component with the saturation component, the weighted quantized hue component and the saturation component are subjected to a statistical operation in order to extract feature vectors. Accordingly, the method and system provide overall segmentation results that are very close to human interpretation.

## Claims

### Claim 1

A method of extracting a perceptual feature set, comprising using a processor to perform the steps of:
converting an input image to obtain a hue component and a saturation component;
quantizing the hue component into a plurality of quantum values;
weighting the quantized hue component with the saturation component; and
subjecting the weighted quantized hue component and the saturation component to a statistical operation in order to extract feature vectors.

### Claim 2 (depends on 1)

The method of claim 1, wherein the input image is converted into a representation in HSI (hue saturation/intensity) color space.

### Claim 3 (depends on 1)

The method of claim 1, wherein the hue component is quantized into six quantum values.

### Claim 4 (depends on 3)

The method of claim 3, wherein the six quantum values are red, yellow, green, cyan, blue and magenta.

### Claim 5 (depends on 1)

The method of claim 1, wherein the quantized hue component is multiplied by the saturation component.

### Claim 6 (depends on 1)

The method of claim 1, wherein a block (B) is centered at a current pixel (p) whose feature is under extraction; and within the block, the feature vectors including chromatic and achromatic feature vectors Hcolor and Hgray are obtained using the saturation components (Sp) with respect to each respective color as a histogram and expressed as
H
color
=
∑
p
∈
color
,
p
∈
B
⁢
S
p
⁢
⁢
color
∈
{
red
,
yellow
,
green
,
cyan
,
blue
,
magenta
}
H
gray
=
∑
p
∈
B
⁢
(
1
-
S
p
)
.

### Claim 7 (depends on 1)

The method of claim 1, further comprising a step of classifying the extracted feature vectors.

### Claim 8 (depends on 7)

The method of claim 7, wherein the extracted feature vectors are classified using a clustering technique.

### Claim 9

A system for extracting a perceptual feature set, comprising:
a conversion unit configured to convert an input image to obtain a hue component and a saturation component;
a quantization unit configured to quantize the hue component into a plurality of quantum values;
a weighting unit configured to weigh the quantized hue component with the saturation component; and
a statistical unit configured to subject the weighted quantized hue component and the saturation component to a statistical operation in order to extract feature vectors.

### Claim 10 (depends on 9)

The system of claim 9, wherein the conversion unit is configured to convert the input image into a representation in HSI (hue/saturation/intensity) color space.

### Claim 11 (depends on 9)

The system of claim 9, wherein the quantization unit is adapted to quantize the hue component into six quantum values.

### Claim 12 (depends on 11)

The system of claim 11, wherein the six quantum values are red, yellow, green, cyan, blue and magenta.

### Claim 13 (depends on 9)

The system of claim 9, wherein the weighing unit is configured to multiply the quantized hue by the saturation component.

### Claim 14 (depends on 9)

The system of claim 9, wherein:
a block (B) is centered at a current pixel (p) whose feature is under extraction;
the saturation unit is configured to obtain from within the block the feature vectors including chromatic and achromatic feature vectors Hcolor and Hgray; and
the feature values are obtained using the saturation components(Sp) with respect to each respective color as a histogram and expressed as
H
color
=
∑
p
∈
color
,
p
∈
B
⁢
S
p
⁢
⁢
color
∈
{
red
,
yellow
,
green
,
cyan
,
blue
,
magenta
}
H
gray
=
∑
p
∈
B
⁢
(
1
-
S
p
)
.

### Claim 15 (depends on 9)

The system of claim 9, further comprising a classifying unit configured to classify the extracted feature vectors.

### Claim 16 (depends on 15)

The system of claim 15, wherein the classifying unit is configured to use a clustering technique to classify the extracted feature vectors.