Learning device, learning method, learning model, detection device and grasping system

Patent metadata

Field	Value
Publication number	US-11034018-B2
Application number	US-201916698177-A
Country	US
Kind code	B2
Filing date	Nov 27, 2019
Priority date	May 31, 2017
Publication date	Jun 15, 2021
Grant date	Jun 15, 2021

Abstract

An estimation device includes a memory and at least one processor. The at least one processor is configured to acquire information regarding a target object. The at least one processor is configured to estimate information regarding a location and a posture of a gripper relating to where the gripper is able to grasp the target object. The estimation is based on an output of a neural model having as an input the information regarding the target object. The estimated information regarding the posture includes information capable of expressing a rotation angle around a plurality of axes.

First claim

The invention claimed is:

1. An estimation device, comprising
at least one memory; and
at least one processor configured to:
acquire information regarding a target object, and
estimate information regarding a location and a posture of a gripper relating to where the gripper is able to grasp the target object, wherein the estimation is based on an output of a neural network model having as an input the information regarding the target object,

wherein the estimated information regarding the posture includes information capable of expressing a rotation angle around a plurality of axes.

2. The estimation device according to claim 1 , wherein
the information capable of expressing the rotation angle around the plurality of axes includes information of a three-dimensional posture angle of the gripper.

3. The estimation device according to claim 2 , wherein
the information of the three-dimensional posture angle includes information regarding a roll angle, a pitch angle and a yaw angle of the gripper with respect to a predetermined reference posture.

4. The estimation device according to claim 1 , wherein
the information capable of expressing the rotation angle around the plurality of axes includes information regarding at least one of Euler angles, an argument, or a direction cosine of the gripper with respect to a predetermined reference posture.

5. The estimation device according to claim 1 , wherein
the information regarding the location of the gripper includes information represented by either an orthogonal coordinate system or a cylindrical coordinate system of the gripper with respect to a predetermined reference point.

6. The estimation device according to claim 1 , wherein
the information regarding the location and the posture of the gripper has a six-dimensional or more degree-of-freedom including angle information of the gripper.

7. The estimation device according to claim 1 , wherein
the at least one processor is configured to estimate information regarding a plurality of classifications of the posture of the gripper.

8. The estimation device according to claim 7 , wherein
the at least one processor is configured to obtain the plurality of classifications by clustering information regarding a posture included in learning data.

9. A grasping system, comprising:
a gripper configured to grasp a target object;
a robot configured to support the gripper; and
a controller configured to control the robot, wherein the controller is configured to control the robot based on the estimated information estimated by the estimation device according to claim 1 .

10. The grasping system according to claim 9 , wherein
the gripper comprises a camera configured to acquire image information of the target object.

11. A learning device, comprising
at least one memory, and
at least one processor configured to learn a learning model which is represented by a neural network model, the learning model having as an input information regarding a target object and as an output information regarding a location and a posture of a gripper relating to where the gripper is able to grasp the target object, wherein
the output information regarding the posture includes information capable of expressing a rotation angle around a plurality of axes.

12. The learning device according to claim 11 , wherein
the information capable of expressing the rotation angle around the plurality of axes includes information of a three-dimensional posture angle of the gripper.

13. The learning device according to claim 12 , wherein
the information of the three-dimensional posture angle includes information regarding a roll angle, a pitch angle and a yaw angle of the gripper with respect to a predetermined reference posture.

14. The learning device according to claim 11 , wherein
the information capable of expressing the rotation angle around the plurality of axes includes information regarding at least one of Euler angles, an argument, or a direction cosine of the gripper with respect to a predetermined reference posture.

15. The learning device according to claim 11 , wherein
the information regarding the location of the gripper includes information represented by either an orthogonal coordinate system or a cylindrical coordinate system of the gripper with respect to a predetermined reference point.

16. The learning device according to claim 11 , wherein
the information regarding the location and the posture of the gripper has a six-dimensional or more degree-of-freedom including angle information of the gripper.

17. The learning device according to claim 11 , wherein
the learning model outputs information regarding a plurality of classifications of the posture of the gripper as the information regarding the posture of the gripper.

18. The learning device according to claim 17 , wherein
the at least one processor is configured to obtain the plurality of classifications by clustering the information regarding the posture included in learning data.

19. An estimation method, comprising:
acquiring, by at least one processor, information regarding a target object; and
estimating, by the at least one processor, information regarding a location and a posture of a gripper relating to where the gripper is able to grasp the target object, based on an output of a neural network model having as an input the information regarding the target object,
wherein the estimated information regarding the posture includes information capable of expressing a rotation angle around a plurality of axes.

20. A learning method, comprising:
learning, by at least one processor, a learning model which is represented by a neural network model, the learning model having as an input information regarding a target object and as an output information regarding a location and a posture of a gripper relating to where the gripper is able to grasp the target object,
wherein the output information regarding the posture includes information capable of expressing a rotation angle around a plurality of axes.

Assignees

• Preferred Networks Inc

Inventors

• Kusano Hitoshi
• KUME Ayaka
• Matsumoto Eiichi

Classifications

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title

• G06N3/09

  Supervised learning · CPC title

• G06T2207/20081

  Training; Learning · CPC title

• G06T7/00

  Image analysis · CPC title

• B25J13/00

  Controls for manipulators (programme controls B25J9/16) · CPC title