# US 9,519,543 B2 · Method and apparatus for image search using feature point

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

- Number: US 9,519,543 B2
- Document ID: US9519543B2
- Status: Granted
- Kind: B2
- Application number: 14/628,959
- Filed: 2015-02-23
- Published: 2015-06-18
- Granted: 2016-12-13
- Earliest claimed: 2010-11-15
- Assignees: Samsung Electronics Co., Ltd.
- Applicants: Samsung Electronics Co., Ltd.
- Inventors: Sang-Wook Kim; Seon Min Rhee; Yong Beom Lee
- Web page: https://patentgrep.com/patent/US9519543B2

## CPC

- G06F 11/1402: {Saving, restoring, recovering or retrying}
- G06F 16/1744: {using compression, e.g. sparse files}
- G06F 16/583: using metadata automatically derived from the content
- G06V 10/462: {Salient features, e.g. scale invariant feature transforms [SIFT]}
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING

## Abstract

A method and an apparatus search an image using a feature point. The image search method extracts at least one feature point from an image and describes the extracted at least one feature point in stages, thereby generating a hierarchical feature point descriptor. In addition, the method may search for information matching the feature point descriptor from a local database (DB) included in a terminal or a remote DB included in a server.

## Claims

### Claim 1

A server performing an image searching method, the server comprising:
a processor comprising:
a descriptor reception unit configured to receive a feature point descriptor in a compressed state from a terminal connected via a network;
a feature point recovery unit configured to recover at least one feature point by decompressing the compressed feature point descriptor; and
an information search unit configured to search for information matching the at least one feature point in a remote database (DB) based on the recovered at least one feature point,
wherein the terminal is configured to calculate an error rate of the feature point descriptor, and to determine whether to additionally describe at least one grid cell located around the at least one feature point based on the error rate, and
wherein the feature point descriptor is information that describes the at least one feature point through an orientation and a position of at least one grid cell located around the at least one feature point.

### Claim 2 (depends on 1)

The server of claim 1, wherein the terminal generates a feature point descriptor by describing at least one feature point extracted from an image, in stages, by order of importance.

### Claim 3 (depends on 2)

The server of claim 2, wherein the terminal generates the feature point descriptor by describing grid cells located around the at least one feature point in order of distance value with respect to the at least one feature point.

### Claim 4 (depends on 2)

The server of claim 2, wherein the terminal generates the feature point descriptor by describing grid cells located around the at least one feature point in order of orientation-related statistics.

### Claim 5 (depends on 1)

The server of claim 1, wherein the feature point recovery unit decompresses the feature point descriptor in order of grid cells constituting the feature point descriptor, and further decompresses the grid cells according to an error rate of a decompression result.

### Claim 6

An image searching method performed by a server, the method comprising:
receiving, by the server, a feature point descriptor in a compressed state from a terminal connected via a network;
recovering, by the server, at least one feature point by decompressing the compressed feature point descriptor; and
searching, by the server, for information matching the at least one feature point from a remote (DB) based on the recovered at least one feature point,
wherein the terminal calculates an error rate of the feature point descriptor, and determines whether to additionally describe at least one grid cell located around the at least one feature point based on the error rate, and
wherein the feature point descriptor is information that describes the at least one feature point through an orientation and a position of at least one grid cell located around the at least one feature point.

### Claim 7 (depends on 6)

The image searching method of claim 6, wherein the terminal generates the feature point descriptor by describing the at least one feature point extracted from an image, in stages, by order of importance.

### Claim 8 (depends on 6)

The image searching method of claim 6, wherein the terminal generates the feature point descriptor by describing grid cells located around the at least one feature point in order of distance value with respect to the at least one feature point.

### Claim 9 (depends on 7)

The image searching method of claim 7, wherein the terminal generates the feature point descriptor by describing grid cells located around the at least one feature point in order of orientation-related statistic.

### Claim 10 (depends on 7)

The image searching method of claim 7, wherein the recovering of the at least one feature point decompresses the feature point descriptor in order of grid cells constituting the feature point descriptor, and further decompresses the grid cells according to the error rate of a decompression result.

### Claim 11 (depends on 6)

A non-transitory computer-readable storage medium storing instructions that, when executed by a processor, cause the processor to perform the method of claim 6.