# US 2018/0322364 A1 · Information processing apparatus, image processing method, and computer-readable recording medium recording image processing program

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

- Number: US 2018/0322364 A1
- Document ID: US20180322364A1
- Status: Published application
- Kind: A1
- Application number: 16/034,969
- Filed: 2018-07-13
- Published: 2018-11-08
- Earliest claimed: 2016-01-20
- Assignees: Fujitsu Limited
- Applicants: Fujitsu Limited
- Inventors: Tsuyoshi Nagato; Hiroaki Okamoto; Tetsuo Koezuka
- Web page: https://patentgrep.com/patent/US20180322364A1

## CPC

- G06V 10/776: Validation; Performance evaluation
- G06F 18/2111: by using evolutionary computational techniques, e.g. genetic algorithms
- G06F 18/217: {Validation; Performance evaluation; Active pattern learning techniques}
- G06N 3/126: Evolutionary algorithms, e.g. genetic algorithms or genetic programming
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

An information processing apparatus includes: a memory configured to store an image processing program having a tree structure in which a partial program is incorporated in each of a plurality of nodes; and a processor configured to performs, based on the image processing program, operations of; calculating a feature amount based on a processing result in each intermediate node excluding a terminal node among the plurality of nodes when executing image processing on a captured image which is captured by an imaging device; and calculating a performance evaluation value of the image processing program based on a variation amount of the feature amount in accordance with elapse of time.

## Claims

### Claim 1

An information processing apparatus comprising:
a memory configured to store an image processing program having a tree structure in which a partial program is incorporated in each of a plurality of nodes; and
a processor configured to performs, based on the image processing program, operations of;
calculating a feature amount based on a processing result in each intermediate node excluding a terminal node among the plurality of nodes when executing image processing on a captured image which is captured by an imaging device; and
calculating a performance evaluation value of the image processing program based on a variation amount of the feature amount in accordance with elapse of time.

### Claim 2 (depends on 1)

The image processing device according to claim 1, wherein the processor determines whether the image processing program is to be regenerated by genetic programming, based on the performance evaluation value.

### Claim 3 (depends on 2)

The image processing device according to claim 2, wherein the processor:
performs initial image processing on an input image which is used in regeneration of the image processing program by the genetic programming;
calculates a node feature amount for the each intermediate node based on the processing result in the each intermediate node; and
calculates a threshold value based on an index representing a variation degree of the node feature amount.

### Claim 4 (depends on 3)

The image processing device according to claim 3, wherein a determination of whether the image processing program is to be regenerated by genetic programming is performed based on comparison between the performance evaluation value and the index.

### Claim 5 (depends on 1)

The image processing device according to claim 1, wherein the processor:
performs initial image processing on an input image which is used in regeneration of the image processing program by genetic programming;
calculates an initial feature amount based on the processing result in the each intermediate node; and
calculates a distance between the feature amount and the initial feature amount as the performance evaluation value, in a calculation of the performance evaluation value.

### Claim 6 (depends on 1)

The image processing device according to claim 1, wherein in a calculation of the performance evaluation value, an index representing a variation degree of the feature amount which is calculated in each of the last multiple times of execution of the image processing is calculated as the performance evaluation value.

### Claim 7 (depends on 1)

The image processing device according to claim 1, wherein in a calculation of the feature amount, a node feature amount is calculated for the each intermediate node by a calculation method which differs depending on a kind of the partial program which is incorporated in the each intermediate node, and the feature amount is calculated based on the node feature amount.

### Claim 8

An information processing method comprising:
calculating, by a computer, based on an image processing program having a tree structure in which a partial program is incorporated in each of a plurality of nodes, a feature amount based on a processing result in each intermediate node excluding a terminal node among the plurality of nodes when executing image processing on a captured image which is captured by an imaging device; and
calculating a performance evaluation value of the image processing program based on a variation amount of the feature amount in accordance with elapse of time.

### Claim 9 (depends on 8)

The image processing method according to claim 8, further comprising:
determining whether the image processing program is to be regenerated by genetic programming, based on the performance evaluation value.

### Claim 10 (depends on 9)

The image processing method according to claim 9, further comprising:
performing initial image processing on an input image which is used in regeneration of the image processing program by the genetic programming;
calculating a node feature amount for the each intermediate node based on the processing result in the each intermediate node; and
calculating a threshold value based on an index representing a variation degree of the node feature amount.

### Claim 11 (depends on 10)

The image processing method according to claim 10, wherein, the determining whether the image processing program is to be regenerated by genetic programming is performed based on comparison between the performance evaluation value and the index.

### Claim 12 (depends on 8)

The image processing method according to claim 8, further comprising:
performing initial image processing on an input image which is used in regeneration of the image processing program by genetic programming;
calculating an initial feature amount based on the processing result in the each intermediate node; and
calculating a distance between the feature amount and the initial feature amount as the performance evaluation value, in a calculation of the performance evaluation value.

### Claim 13 (depends on 8)

The image processing method according to claim 8, wherein in the calculating the performance evaluation value, an index representing a variation degree of the feature amount which is calculated in each of the last multiple times of execution of the image processing is calculated as the performance evaluation value.

### Claim 14 (depends on 8)

The image processing device according to claim 8, wherein in the calculating the feature amount, a node feature amount is calculated for the each intermediate node by a calculation method which differs depending on a kind of the partial program which is incorporated in the each intermediate node, and the feature amount is calculated based on the node feature amount.

### Claim 15

A non-transitory computer-readable recording medium recording an image processing program which causes a computer to perform a process, the process comprising:
calculating based on the image processing program having a tree structure in which a partial program is incorporated in each of a plurality of nodes, a feature amount based on a processing result in each intermediate node excluding a terminal node among the plurality of nodes when executing image processing on a captured image which is captured by an imaging device; and
calculating a performance evaluation value of the image processing program based on a variation amount of the feature amount in accordance with elapse of time.

### Claim 16 (depends on 15)

The non-transitory computer-readable recording medium according to claim 15, further comprising:
determining whether the image processing program is to be regenerated by genetic programming, based on the performance evaluation value.

### Claim 17 (depends on 16)

The non-transitory computer-readable recording medium according to claim 16, further comprising:
performing initial image processing on an input image which is used in regeneration of the image processing program by the genetic programming;
calculating a node feature amount for the each intermediate node based on the processing result in the each intermediate node; and
calculating a threshold value based on an index representing a variation degree of the node feature amount.

### Claim 18 (depends on 17)

The non-transitory computer-readable recording medium according to claim 17, wherein, the determining whether the image processing program is to be regenerated by genetic programming is performed based on comparison between the performance evaluation value and the index.

### Claim 19 (depends on 15)

The non-transitory computer-readable recording medium according to claim 15, further comprising:
performing initial image processing on an input image which is used in regeneration of the image processing program by genetic programming;
calculating an initial feature amount based on the processing result in the each intermediate node; and
calculating a distance between the feature amount and the initial feature amount as the performance evaluation value, in a calculation of the performance evaluation value.

### Claim 20 (depends on 15)

The non-transitory computer-readable recording medium according to claim 15, wherein in the calculating the performance evaluation value, an index representing a variation degree of the feature amount which is calculated in each of the last multiple times of execution of the image processing is calculated as the performance evaluation value.
