Method, device, system and computer-program product for setting lighting condition and storage medium

Patent metadata

Field	Value
Publication number	US-2020410270-A1
Application number	US-201816970376-A
Country	US
Kind code	A1
Filing date	Mar 6, 2018
Priority date	Mar 6, 2018
Publication date	Dec 31, 2020
Grant date	—

Abstract

The present disclosure relates to a method, device, system and computer-program product for setting a lighting condition when an object is checked and a storage medium. The method includes that: the object is lighted by light sources capable of changing lighting parameters, and the object is captured by an image sensor in such lighting parameters to obtain captured images, wherein the object has known label data; and a part of or all of the captured images and the corresponding label data of the object are applied to learning of a machine learning model, and the lighting condition and the check algorithm parameters of the machine learning model is set simultaneously by optimizing both the lighting parameters and the check algorithm parameters, on the basis of a comparison result between an estimation result of the machine learning model and the label data. Therefore, operations are simplified.

First claim

1 . A method for setting a lighting condition when an object is checked, wherein, a check module comprises a machine learning model is used to check the object, the machine learning model being generated by using learning data comprising images, the method is characterized by comprising:
lighting the object by light sources capable of changing lighting parameters specifying the lighting condition when the object is captured, and capturing the object by an image sensor in a plurality of lighting parameters, so as to obtain captured images corresponding to the plurality of lighting parameters, wherein, the object has known label data; and applying a part of or all of the captured images corresponding to the plurality of lighting parameters and the corresponding label data of the object to the learning of the machine learning model, and setting both the lighting condition and the check algorithm parameters of the machine learning model by simultaneously optimizing both the lighting parameters and the check algorithm parameters, on the basis of the comparison result between the estimation result of the machine learning model and the label data of the object.

2 . A method for setting a lighting condition when an object is checked, wherein, a check module comprises a machine learning model is used to check the object, the machine learning model being generated by using learning data comprising images, the method is characterized by comprising:
lighting the object by light sources capable of changing lighting parameters specifying the lighting condition when the object is captured, and capturing the object by an image sensor in a plurality of lighting parameters, so as to obtain a plurality of captured images corresponding to the plurality of lighting parameters, wherein, the object has known label data; and applying a part of or all of the captured images corresponding to the plurality of lighting parameters to the machine learning model which has already performed learning, and setting the lighting condition by optimizing only a selection of the predetermined lighting parameters on the basis of the comparison result between the estimation result of the machine learning model and the label data of the object.

3 . The method of claim 2 , wherein, applying a part of or all of the captured images corresponding to the plurality of lighting parameters to the machine learning model which has performed learning comprises:
applying learning data comprising the captured images and the corresponding label data of the object to the additional learning of the machine learning model, so as to update a part of or all of the check algorithm parameters of the machine learning model, wherein, the label data represents checked features of the object; and optimizing both the selection of the lighting parameters and a part of or all of the check algorithm parameters of the machine learning model so as to make the estimation result of the machine learning model be consistent with the label data.

4 . The method of claim 2 , wherein,
when the lighting condition is set, the number of the captured images which are applied to the machine learning model which has already performed learning in order to find the optimal lighting condition is less than the number of the captured images which are applied to the learning of the machine learning model.

5 . The method of claim 2 , wherein,
the lighting parameters comprises luminous positions and luminous intensity of the light sources.

6 . The method of claim 2 , wherein, setting the lighting condition when the check module is used to check the object comprises:
selecting the lighting parameters that minimize the loss function representing the comparison result, wherein, the lighting parameters are variables of the loss function, wherein, selecting the lighting parameters comprises: selecting the lighting parameters that minimize the loss average of the loss function for the lighting parameters of a predetermined range.

7 . A device for setting a lighting condition when an object is checked, wherein, a check module comprises a machine learning model is used to check the object, the machine learning model being generated by using learning data comprising images, the device is characterized by comprising:
an acquisition unit acquiring captured images regarding the object, wherein, the object being lighted by light sources capable of changing lighting parameters specifying the lighting condition when the object is captured, and the object being captured by an image sensor in a plurality of lighting parameters, so as to obtain the captured images corresponding to the plurality of lighting parameters, wherein, the object has known label data; and a setting unit, applying a part of or all of the captured images corresponding to the plurality of lighting parameters and the corresponding label data of the object to the learning of the machine learning model, and setting both the lighting condition and the check algorithm parameters of the machine learning model by simultaneously optimizing both the lighting parameters and the check algorithm parameters, on the basis of the comparison result between the estimation result of the machine learning model and the label data of the object.

8 . A device for setting a lighting condition when an object is checked, wherein, a check module comprises a machine learning model is used to check the object, the machine learning model being generated by using learning data comprising images, the device is characterized by comprising:
an acquisition unit acquiring captured images regarding the object, wherein, the object being lighted by light sources capable of changing lighting parameters specifying the lighting condition when the object is captured, and the object being captured by an image sensor in a plurality of lighting parameters, so as to obtain the captured images corresponding to the plurality of lighting parameters, wherein, the object has known label data; and a setting unit, applying a part of or all of the captured images corresponding to the plurality of lighting parameters to the machine learning model which has already performed learning, and setting the lighting condition by optimizing only a selection of the predetermined lighting parameters on the basis of the comparison result between the estimation result of the machine learning model and the label data of the object.

9 . The device of claim 8 , wherein, the setting unit:
applying learning data comprising the captured images and the corresponding label data of the object to the additional learning of the machine learning model, so as to update a part of or all of the check algorithm parameters of the machine learning model, wherein, the label data represents checked features of the object; and optimizing both the selection of the lighting parameters and a part of or all of the check algorithm parameters of the machine learning model so as to make the estimation result of the machine learning model be consistent with the label data.

10 . The device of claim 8 , wherein,
when the lighting condition is set, the number of the captured images which are applied to the machine learning model which has already performed learning in order to find the optimal lighting condition is less than the number of the captured images which are applied to the learning of the machine learning model.

11 . The device of claim 8 , wherein,
the lighting parameters comprises luminous positions and luminous intensity of the light sources.

12 . The device of claim 8 , wherein, the setting unit:
selecting the lighting parameters that minimize the loss function representing the com

Assignees

• Omron Tateisi Electronics Co

Inventors

• Naruse Yosuke
• Kurita Masashi

Classifications

• G01N21/8806 Primary

  Specially adapted optical and illumination features · CPC title

• G06N3/09

  Supervised learning · CPC title

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title

• G06V10/141

  Control of illumination · CPC title

• G06T7/001

  using an image reference approach · CPC title