# US 11,113,580 B2 · Image classification system and method

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

- Number: US 11,113,580 B2
- Document ID: US11113580B2
- Status: Granted
- Kind: B2
- Application number: 16/731,001
- Filed: 2019-12-30
- Published: 2021-07-01
- Granted: 2021-09-07
- Earliest claimed: 2019-12-30
- Assignees: Industrial Technology Research Institute
- Applicants: Industrial Technology Research Institute
- Inventors: Jiung-Yao Huang; Hsin-Lung Wu; Juin-Ming Lu
- Web page: https://patentgrep.com/patent/US11113580B2

## CPC

- G06V 10/454: {Integrating the filters into a hierarchical structure, e.g. convolutional neural networks [CNN]}
- G06F 18/2113: by ranking or filtering the set of features, e.g. using a measure of variance or of feature cross-correlation
- G06F 18/23: Clustering techniques
- G06F 18/23213: with fixed number of clusters, e.g. K-means clustering
- G06F 18/24137: {Distances to cluster centroïds}
- G06T 3/40: Scaling of whole images or parts thereof, e.g. expanding or contracting
- G06T 9/00: Image coding (bandwidth or redundancy reduction for static pictures H04N1/41; coding or decoding of static colour picture signals H04N1/64; methods or arrangements for coding, decoding, compressing or decompressing digital video signals H04N19/00)
- G06V 10/763: {Non-hierarchical techniques, e.g. based on statistics of modelling distributions}
- G06V 10/82: using neural networks
- G06V 40/1365: {Matching; Classification}
- G06V 40/172: {Classification, e.g. identification}
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL

## Abstract

An image classification system includes a storage device, a computing device and a first processing device. The storage device stores a plurality of pseudo-centroid datasets, wherein the pseudo-centroid datasets correspond to a plurality of units of first image dataset, and the number of pseudo-centroid data points of each of the pseudo-centroid datasets is much smaller than the number of data points of each of the units of first image dataset. The computing device receives the second image data and computes a plurality of feature values of the second image data. The first processing device receives the feature values and the pseudo-centroid datasets, and compares the feature values with the pseudo-centroid data points to identify and classify the second image data.

## Claims

### Claim 1

An image classification system, comprising:
a non-transitory storage device, configured to store a plurality of pseudo-centroid datasets, wherein the pseudo-centroid datasets correspond to a plurality of units of first image dataset, and the number of pseudo-centroid data points of each of the pseudo-centroid datasets is much smaller than the number of data points of each of the units of first image dataset;
a computing device, configured to receive second image data and compute a plurality of feature values of the second image data; and
a first processing device, configured to receive the feature values and the pseudo-centroid datasets, and compare the feature values with the pseudo-centroid data points to identify and classify the second image data.

### Claim 2 (depends on 1)

The image classification system as claimed in claim 1, wherein the first processing device further defines a threshold range of each of the pseudo-centroid data points, the processing device determines whether the feature values fall into the threshold ranges, when the feature values fall into the threshold ranges, the first processing device computes weights of the feature values falling into the threshold ranges, sums up the weights of the feature values falling into the threshold ranges, and selects the pseudo-centroid dataset corresponding to a highest weight as a category of the second image data, and when determining that the feature values do not fall into the threshold ranges, the first processing device determines that the second image data is outlier data.

### Claim 3 (depends on 1)

The image classification system as claimed in claim 1, wherein the computing device comprises:
an image compressing device, configured to compress the second image data.

### Claim 4 (depends on 3)

The image classification system as claimed in claim 3, wherein the image compressing device compresses the second image data through a convolution operation and a max pooling operation.

### Claim 5 (depends on 1)

The image classification system as claimed in claim 1, wherein the computing device further computes a plurality of feature values of the units of first image dataset, and the image classification system further comprises:
a second processing device, configured to receive the feature values of the units of first image dataset, compute a correlation of the feature values of the units of first image dataset, and cluster the data points of the units of first image dataset to generate a plurality of clusters;
wherein the first processing device receives the plurality of clusters, computes a centroid point of each of the clusters, defines a threshold range of the centroid point, and selects the data points of each of the clusters falling into the threshold range of the centroid point as the pseudo-centroid data points, wherein the pseudo-centroid data points are comprised in the corresponding pseudo-centroid dataset.

### Claim 6 (depends on 5)

The image classification system as claimed in claim 5, wherein the second processing device computes the correlation of the feature values of the units of first image dataset, and clusters the data points of the units of first image dataset to generate the plurality of clusters through a similarity matrix algorithm, a Laplacian matrix algorithm, and a K-means clustering algorithm.

### Claim 7 (depends on 1)

The image classification system as claimed in claim 1, wherein the units of first image dataset are different from each other.

### Claim 8 (depends on 1)

The image classification system as claimed in claim 1, wherein the units of first image dataset and the second image data comprise a human face or a fingerprint.

### Claim 9

An image classification method, comprising:
storing a plurality of pseudo-centroid datasets, wherein the pseudo-centroid datasets correspond to a plurality of units of first image dataset, and the number of pseudo-centroid data points of each of the pseudo-centroid datasets is much smaller than the number of data points of each of the units of first image dataset;
receiving second image data and computing a plurality of feature values of the second image data; and
receiving the feature values and the pseudo-centroid datasets, and comparing the feature values with the pseudo-centroid data points to identify and classify the second image data.

### Claim 10 (depends on 9)

The image classification method as claimed in claim 9, wherein the step of receiving the feature values and the pseudo-centroid datasets and comparing the feature values with the pseudo-centroid data points to identify and classify the second image data comprises:
defining a threshold range of each of the pseudo-centroid data points;
determining whether the feature values fall into the threshold ranges;
when determining that the feature values fall into the threshold ranges, computing weights of the feature values falling into the threshold ranges, summing up the weights of the feature values falling into the threshold ranges, and selecting the pseudo-centroid dataset corresponding to a highest weight as a category of the second image data; and
when determining that the feature values do not fall into the threshold ranges, determining that the second image data is outlier data.

### Claim 11 (depends on 9)

The image classification method as claimed in claim 9, wherein the step of receiving the second image data and computing the plurality of feature values of the second image data comprises:
compressing the second image data, and computing the plurality of feature values of the second image data.

### Claim 12 (depends on 11)

The image classification method as claimed in claim 11, wherein the second image data is compressed through a convolution operation and a max pooling operation.

### Claim 13 (depends on 9)

The image classification method as claimed in claim 9, further comprising:
receiving the units of first image dataset, and computing a plurality of feature values of the units of first image dataset;
receiving the feature values of the units of first image dataset, computing a correlation of the feature values of the units of first image dataset, and clustering the data points of the units of first image dataset to generate a plurality of clusters;
receiving the plurality of clusters, and computing a centroid point of each of the clusters;
defining a threshold range for the centroid point; and
selecting the data points of each of the clusters falling into the threshold range of the centroid point as the pseudo-centroid data points, wherein the pseudo-centroid data points are comprised in the corresponding pseudo-centroid dataset.

### Claim 14 (depends on 13)

The image classification method as claimed in claim 13, wherein the method includes computing the correlation of the feature values of the units of first image dataset, and clustering the data points of the units of first image dataset to generate the plurality of clusters through a similarity matrix algorithm, a Laplacian matrix algorithm, and a K-means clustering algorithm.

### Claim 15 (depends on 9)

The image classification method as claimed in claim 9, wherein the units of first image dataset are different from each other.

### Claim 16 (depends on 9)

The image classification method as claimed in claim 9, wherein the units of first image dataset and the second image data comprise a human face or a fingerprint.