# US 2020/0393390 A1 · Sorting support apparatus, sorting support system, sorting support method, and program

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

- Number: US 2020/0393390 A1
- Document ID: US20200393390A1
- Status: Published application
- Kind: A1
- Application number: 16/971,715
- Filed: 2019-03-28
- Published: 2020-12-17
- Earliest claimed: 2018-03-29
- Assignees: NEC Corporation
- Applicants: NEC Corporation
- Inventors: Masahiro KAJI; Kazuyuki IKEMURA; Shinichi YOSHITSUNE; Kaori TAKEUCHI
- Web page: https://patentgrep.com/patent/US20200393390A1

## CPC

- B07C 5/346: according to radioactive properties {(material testing G01N23/00; radiation measurement G01T)}
- G01N 23/10: the material being confined in a container, e.g. in a luggage X-ray scanners
- G01N 23/04: and forming images of the material
- G01N 23/083: the radiation being X-rays
- G06N 5/04: Inference or reasoning models
- G06N 20/00: Machine learning
- G06T 7/00: Image analysis
- G06V 20/52: Surveillance or monitoring of activities, e.g. for recognising suspicious objects (recognising microscopic objects G06V20/69)
- G01N 2223/04: and measuring absorption
- G01N 2223/401: image processing
- G01N 2223/639: material in a container
- B07C: POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING (specially adapted for a specific purpose covered by another class, see the relevant place, e.g. A43D33/06, B23Q7/12)
- G01N: INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES (measuring or testing apparatus or processes other than immunoassay, involving enzymes or microorganisms C12M, C12Q)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING

## Abstract

A sorting support apparatus is provided with: an input part that inputs a transmission image obtained by radiating an inspection target with electromagnetic waves; a storage part that stores a plurality of learning models optimized respectively for at least one article and being associated with an assumed usage condition; and a determination part that selects one of the learning models based on a specified usage condition and uses the learning model to determine whether or not the one or more articles is contained in the inspection target.

## Claims

### Claim 1

A sorting support apparatus comprising:
an input part configured to input a transmission image obtained by radiating an inspection target with electromagnetic waves;
a storage part configured to store a plurality of learning models optimized respectively for at least one article and being associated with an assumed usage condition; and
a determination part configured to select one of the learning models based on a specified usage condition and uses the learning model to determine whether or not the one or more articles is contained in the inspection target.

### Claim 2 (depends on 1)

The sorting support apparatus according to claim 1, wherein the learning model is created in accordance with a trend of handled goods at a location where the sorting support apparatus is disposed.

### Claim 3 (depends on 1)

The sorting support apparatus according to claim 1, wherein the learning model is created in accordance with a trend of handled goods in a time-period in which sorting is performed.

### Claim 4 (depends on 1)

The sorting support apparatus according to claim 1, wherein the learning model is created in accordance with a trend of handled goods according to sender location.

### Claim 5 (depends on 1)

The sorting support apparatus according to claim 1, wherein it is possible to change a threshold for determining, in the determination part, whether or not the at least one article is included.

### Claim 6 (depends on 1)

A sorting support system wherein the sorting support apparatus of claim 1 is disposed at multiple stages to determine in a stepwise manner whether or not the at least one article is included, using different learning models.

### Claim 7 (depends on 6)

The sorting support system according to claim 6, configured so that
a sorting support apparatus at a first stage uses a learning model optimized for sorting paper and non-paper articles, and
sorting support apparatuses at second and following stages use learning models optimized for further sorting the non-paper articles.

### Claim 8

A sorting support method, wherein a sorting support apparatus that comprises
an input part configured to input a transmission image obtained by radiating an inspection target with electromagnetic waves, and
a storage part configured to store a plurality of learning models optimized respectively for at least one article and being associated with an assumed usage condition:
selects one of the learning models based on a specified usage condition; and
determines, by using the learning model, whether or not the one or more articles is included in the inspection target.

### Claim 9

(canceled)

### Claim 10 (depends on 2)

The sorting support apparatus according to claim 2, wherein the learning model is created in accordance with a trend of handled goods in a time-period in which sorting is performed.

### Claim 11 (depends on 2)

The sorting support apparatus according to claim 2, wherein the learning model is created in accordance with a trend of handled goods according to sender location.

### Claim 12 (depends on 3)

The sorting support apparatus according to claim 3, wherein the learning model is created in accordance with a trend of handled goods according to sender location.

### Claim 13 (depends on 2)

The sorting support apparatus according to claim 2, wherein it is possible to change a threshold for determining, in the determination part, whether or not the at least one article is included.

### Claim 14 (depends on 3)

The sorting support apparatus according to claim 3, wherein it is possible to change a threshold for determining, in the determination part, whether or not the at least one article is included.

### Claim 15 (depends on 4)

The sorting support apparatus according to claim 4, wherein it is possible to change a threshold for determining, in the determination part, whether or not the at least one article is included.

### Claim 16 (depends on 2)

A sorting support system wherein the sorting support apparatus of claim 2 is disposed at multiple stages to determine in a stepwise manner whether or not the at least one article is included, using different learning models.

### Claim 17 (depends on 3)

A sorting support system wherein the sorting support apparatus of claim 3 is disposed at multiple stages to determine in a stepwise manner whether or not the at least one article is included, using different learning models.

### Claim 18 (depends on 4)

A sorting support system wherein the sorting support apparatus of claim 4 is disposed at multiple stages to determine in a stepwise manner whether or not the at least one article is included, using different learning models.

### Claim 19 (depends on 5)

A sorting support system wherein the sorting support apparatus of claim 5 is disposed at multiple stages to determine in a stepwise manner whether or not the at least one article is included, using different learning models.

### Claim 20 (depends on 16)

The sorting support system according to claim 16, configured so that
a sorting support apparatus at a first stage uses a learning model optimized for sorting paper and non-paper articles, and
sorting support apparatuses at second and following stages use learning models optimized for further sorting the non-paper articles.
