Grain size estimation device, grain size estimation method, grain size estimation program, and grain size estimation system

Patent metadata

Field	Value
Publication number	US-12067703-B2
Application number	US-202017441473-A
Country	US
Kind code	B2
Filing date	Sep 18, 2020
Priority date	Sep 18, 2020
Publication date	Aug 20, 2024
Grant date	Aug 20, 2024

Abstract

According to one embodiment, a grain size estimation device includes an acquisition unit that acquires a captured image of a surface segment of an object inducing metal; and an estimation unit that estimates a grain size of the surface segment of the object indicated in the acquired image, based on a predictive model generated by machine learning using images of metal surfaces and grain sizes in the metal surfaces as training data.

First claim

The invention claimed is:

1. A grain size estimation device comprising:
an acquisition unit that acquires a captured image of a surface segment of an object including metal;
an estimation unit that estimates a grain size of the surface segment of the object indicated in the acquired image, based on a predictive model generated by machine learning using images of metal surfaces and grain sizes on the metal surfaces as training data; and
a determination unit that determines whether or not the object is good based on the estimation result made by the estimation unit.

2. The grain size estimation device according to claim 1
further comprising a training unit
wherein when the grain size estimated by the estimation unit is modified by a user, the training unit updates the predictive model by machine learning using the modified grain size and the captured image associated with the modified grain size as training data.

3. The grain size estimation device according to claim 1
wherein:
the acquisition unit acquires respective captured images of multiple surface segments on the object;
the estimation unit estimates respective grain size of each of the multiple surface segments on the object shown in the captured images; and
an evaluation unit determines that the object is good when an average value of the respective estimated grain sizes exceeds a predetermined threshold value.

4. The grain size estimation device according to claim 1
further comprising a presentation unit that presents a result information including an estimated grain size to a user, and
when the user changes the result information, the grain size estimation device attaches an identifying information to the changed result information that distinguishes the changed result information from the result information that is not changed.

5. The grain size estimation device according to claim 4
further comprising a memory wherein, when a user makes a change to the result information, the memory memorizes the change the user made.

6. The grain size estimation device according to claim 1
wherein:
the acquisition unit acquires respective captured images of multiple surface segments on the object;
the estimation unit estimates respective grain size of each of the multiple surface segments on the object shown in the captured images; and
a result information including estimated grain sizes and a distribution of the grain sizes of each surface segment on the object is displayed on a display device.

7. A grain size estimation method
wherein a computer
acquires a captured image of a surface segment of an object including metal, and
estimates a grain size of the surface segment of the object indicated in the acquired image, based on a predictive model generated by machine learning using images of metal surfaces and grain sizes on the metal surfaces as training data, and
determines whether or not the object is good based on the estimation result made by the estimating.

8. A non-transitory computer readable medium including a grain size estimation program that makes a computer function as:
an acquisition unit that acquires a captured image of a surface segment of an object including metal,
an estimation unit that estimates a grain size of the surface segment of the object indicated in the acquired image, based on a predictive model generated by machine learning using images of metal surfaces and grain sizes on the metal surfaces as training data, and
a determination unit that determines whether or not the object is good based on the estimation result made by the estimation unit.

9. A grain size estimation system comprising:
an image capturing device that obtains a captured image by capturing an image of a surface segment of an object including metal, and
a server that is communicably connected to the image capturing device via a network, wherein the image capturing device comprises
a transmission unit that transmits the captured image to the server, and

wherein the server comprises
an acquisition unit that acquires the captured image from the image capturing device,
an estimation unit that estimates a grain size of the surface segment of the object indicated in the acquired image, based on a predictive model generated by machine learning using images of metal surfaces and grain sizes in the metal surfaces as training data, and
a determination unit that determines whether or not the object is good based on the estimation result made by the estimation unit.

10. A grain size estimation device comprising:
an acquisition unit that acquires a captured image of a surface segment of an object including metal;
an estimation unit that estimates a grain size of the surface segment of the object indicated in the acquired image, based on a predictive model generated by machine learning using images of metal surfaces and grain sizes on the metal surfaces as training data; and
a training unit,
wherein when the grain size estimated by the estimation unit is modified by a user, the training unit updates the predictive model by machine learning using the modified grain size and the captured image associated with the modified grain size as training data.

11. A grain size estimation device comprising:
an acquisition unit that acquires a captured image of a surface segment of an object including metal;
an estimation unit that estimates a grain size of the surface segment of the object indicated in the acquired image, based on a predictive model generated by machine learning using images of metal surfaces and grain sizes on the metal surfaces as training data;
a presentation unit that presents a result information including an estimated grain size to a user; and
when the user changes the result information, the grain size estimation device attaches an identifying information to the changed result information that distinguishes the changed result information from the result information that is not changed.

12. A grain size estimation device comprising:
an acquisition unit that acquires a captured image of a surface segment of an object including metal;
an estimation unit that estimates a grain size of the surface segment of the object indicated in the acquired image, based on a predictive model generated by machine learning using images of metal surfaces and grain sizes on the metal surfaces as training data; and
wherein:
the acquisition unit acquires respective captured images of multiple surface segments on the object;
the estimation unit estimates respective grain size of each of the multiple surface segments on the object shown in the captured images; and
a result information including estimated grain sizes and a distribution of the grain sizes of each surface segment on the object is displayed on a display device.

Assignees

• Toshiba Kk
• Toshiba Digital Solutions Corp

Inventors

• Fukuyama Kota
• Hata Hisanori
• Ikeya Takayuki

Classifications

• G06T2207/30136

  Metal · CPC title

• G06T2207/20081

  Training; Learning · CPC title

• G06T2207/10056

  Microscopic image · CPC title

• G06T7/60

  Analysis of geometric attributes · CPC title

• G01N33/204 Primary

  Structure thereof, e.g. crystal structure · CPC title