WO2021235311A1 - 機械学習装置及び機械学習システム - Google Patents

機械学習装置及び機械学習システム Download PDF

Info

Publication number
WO2021235311A1
WO2021235311A1 PCT/JP2021/018191 JP2021018191W WO2021235311A1 WO 2021235311 A1 WO2021235311 A1 WO 2021235311A1 JP 2021018191 W JP2021018191 W JP 2021018191W WO 2021235311 A1 WO2021235311 A1 WO 2021235311A1
Authority
WO
WIPO (PCT)
Prior art keywords
learning
learning data
lightweight
unit
machine learning
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/JP2021/018191
Other languages
English (en)
French (fr)
Japanese (ja)
Inventor
勇太 並木
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Fanuc Corp
Original Assignee
Fanuc Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Fanuc Corp filed Critical Fanuc Corp
Priority to CN202180035638.7A priority Critical patent/CN115668283A/zh
Priority to JP2022524417A priority patent/JPWO2021235311A1/ja
Priority to US17/998,351 priority patent/US20230186457A1/en
Priority to DE112021002846.4T priority patent/DE112021002846T5/de
Publication of WO2021235311A1 publication Critical patent/WO2021235311A1/ja
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • G06T7/0008—Industrial image inspection checking presence/absence
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
    • G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20081—Training; Learning
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30108—Industrial image inspection
    • G06T2207/30164—Workpiece; Machine component

Definitions

  • the present invention relates to a machine learning device and a machine learning system.
  • machine learning using a learning device such as a deep neural network has been used as a method for detecting and inspecting an object from the features shown in an image.
  • annotations are performed to associate labels such as whether or not there is a defective part in the image and whether or not the detection position is correct with the image data.
  • Annotation is performed by checking the images one by one by a person and visually determining whether or not the object in the image has a defective part.
  • the pair of the image and the label is the training data
  • the set of the training data is the training data set. Then, machine learning is performed by a learner using all or a part of the training data set (see, for example, Patent Documents 1 and 2).
  • the machine learning device includes a machine learning unit that learns an image and learning data including a label for the image, an image processing unit that performs image processing on the image using an image processing program, and the machine learning unit.
  • a lightweight learning data creation unit that cuts out a partial image to be used for training by the unit and creates lightweight learning data including the partial image, and a training data control that stores the lightweight learning data in association with the image processing program.
  • the machine learning unit includes, and the machine learning unit learns the learning data or the lightweight learning data.
  • the machine learning apparatus includes a machine learning unit that learns an image and learning data including a label for the image, an image processing unit that performs image processing on the image using an image processing program, and the machine learning unit.
  • a lightweight learning data creation unit that cuts out a partial image to be used for learning by the unit from the image and creates lightweight learning data including the partial image, a learning model for learning the training data, and the lightweight learning data.
  • the machine learning unit includes a learning data control unit that stores the learning data in association with a lightweight learning model for learning, and the machine learning unit learns the learning data or the lightweight learning data.
  • a machine learning system including a plurality of machine learning devices according to the present disclosure, in which a learning model is shared by machine learning units included in each of the plurality of machine learning devices, and the machine learning included in each of the plurality of machine learning devices.
  • the department learns from the shared learning model.
  • a machine learning system including a plurality of machine learning devices according to the present disclosure, in which lightweight learning data is shared by machine learning units included in each of the plurality of machine learning devices, and the machines included in each of the plurality of machine learning devices.
  • the learning unit performs learning using the shared lightweight learning data.
  • the learning data can be reduced in weight and learning can be performed at high speed.
  • FIG. 1 is a diagram showing an outline of an image processing system 100 to which the machine learning device 10 according to the present embodiment is applied.
  • the image processing system 100 includes an image processing device 1, an object 2, a visual sensor 3, and a workbench 4.
  • the image processing system 100 captures an object 2 arranged on the workbench 4 by the visual sensor 3, and processes the captured image data by the image processing device 1. Further, the image processing device 1 includes a machine learning device 10. The machine learning device 10 uses a learning model to learn a learning data set containing one or more learning data including images and labels.
  • FIG. 2 is a diagram showing an outline of a robot system 200 to which the machine learning device 10 according to the present embodiment is applied.
  • the robot system 200 includes an image processing device 1, an object 2, a visual sensor 3, a workbench 4, a robot 20, and a robot control device 25.
  • a hand or tool is attached to the tip of the arm 21 of the robot 20.
  • the robot 20 performs work such as handling or processing of the object 2 under the control of the robot control device 25.
  • a visual sensor 3 is attached to the tip of the arm 21 of the robot 20.
  • the visual sensor 3 may not be attached to the robot 20, and may be fixedly installed at a predetermined position, for example.
  • the visual sensor 3 takes an image of the object 2 under the control of the image processing device 1.
  • a two-dimensional camera having an image pickup element composed of a CCD (Charge Coupled Device) image sensor and an optical system including a lens may be used, and a stereo camera or the like capable of three-dimensional measurement is used. You may.
  • CCD Charge Coupled Device
  • the robot control device 25 executes a robot program for the robot 20 and controls the operation of the robot 20. At that time, the robot control device 25 corrects the operation of the robot 20 so that the robot 20 performs a predetermined work with respect to the position of the object 2 detected by the image processing device 1.
  • the image processing device 1 includes a machine learning device 10.
  • the machine learning device 10 uses a learning model to learn a learning data set containing one or more learning data including images and labels.
  • FIG. 3 is a diagram showing the configuration of the machine learning device 10.
  • the machine learning device 10 is a device for performing machine learning for the robot 20.
  • the machine learning device 10 includes a control unit 11 and a storage unit 12.
  • the control unit 11 is a processor such as a CPU (Central Processing Unit), and realizes various functions by executing a program stored in the storage unit 12.
  • CPU Central Processing Unit
  • the control unit 11 includes a teaching unit 111, an object detection unit 112, a labeling unit 113, an image processing unit 114, a machine learning unit 115, a lightweight learning data creation unit 116, and a learning data control unit 117.
  • a display control unit 118 is provided.
  • the storage unit 12 stores a ROM (Read Only Memory) for storing an OS (Operating System), an application program, a RAM (Random Access Memory), a hard disk drive for storing various other information, and an SSD (Solid State Drive). It is a device.
  • the storage unit 12 stores various information such as a learning model, learning data, and a robot program.
  • the teaching unit 111 teaches a model pattern that represents the characteristics of the image of the object 2.
  • the object 2 to be taught as a model pattern is arranged in the field of view of the visual sensor 5, and an image of the object 2 is captured. It is desirable that the positional relationship between the visual sensor 3 and the object 2 is the same as when the object 2 is detected.
  • the teaching unit 111 designates a region including the object 2 in the captured image as a rectangular or circular model pattern designation region.
  • the teaching unit 111 extracts edge points as feature points within the range of the model pattern designation area, and obtains physical quantities such as the position of the edge points, the posture (direction of the luminance gradient), and the magnitude of the luminance gradient. Further, the teaching unit 111 defines a model pattern coordinate system in the designated area, and converts the positions and postures of the edge points from the values expressed in the image coordinate system to the values expressed in the model pattern coordinate system. do.
  • the physical quantity of the extracted edge points is stored in the storage unit 12 as feature points constituting the model pattern.
  • edge points are used as feature points, but feature points such as well-known SIFT may be used.
  • a method as disclosed in Japanese Patent Application Laid-Open No. 2017-91079 may be used.
  • the object detection unit 112 detects the image of the object W from one or more input images including the object 2 by using the model pattern. Specifically, first, one or more input images including the image of the object 2 are prepared. Then, the object detection unit 112 detects the image of the object W from each of one or more input images including the object 2 by using the model pattern.
  • Detection parameters are, for example, size range for the model, shear deformation range, detection position range, angle range, matching ratio of model pattern edge to image edge, model pattern edge matching image edge. It may be a threshold value of the distance considered to have been achieved, or a threshold value of the contrast of the edge.
  • the label assigning unit 113 assigns a label (annotation) to the detection result based on the judgment of the detection result of the object 2 by the user. Specifically, the detection result of the object 2 is displayed on the display device 40 connected to the machine learning device 10. The user visually confirms the detection result and assigns a label such as OK or NG to the detection result. When a plurality of objects W are detected from one input image, a plurality of labels are given to one input image.
  • FIG. 4 is a diagram showing an example of assigning a label to the detection result.
  • the labeling unit 113 assigns an NG label to the two images G12 and G17, and assigns an OK label to the six images G11, G13, G14, G15, G16 and G18.
  • the user gives an NG label when the detection result is erroneous detection or defective. Further, the user may give an OK label when the detection result is equal to or more than a predetermined threshold value, and may give an NG label when the detection result is less than the predetermined threshold value. Further, the label automatically given by the machine learning device 10 may be modified by the user. In the above description, the label uses a classification having two classes, OK and NG, but a classification having three or more classes may be used.
  • the image processing unit 114 associates an image with a label for the image, and uses the image and the label as learning data.
  • the data stored as the label may include the data included in the detection result in addition to the OK and NG labels given by the user.
  • the image processing unit 114 since the object is cut out from the input image by using the information such as the position / orientation and the size of the object included in the detection result, it is necessary to store this information as a label. If you cut out the image when creating the training data, you do not need this information.
  • the image processing unit 114 stores an image and a set of training data (learning data set) including labels for the images in the training data storage unit 121.
  • the machine learning unit 115 learns an image and a learning data set including labels for the images.
  • the machine learning unit 115 inputs each pixel value of the image into the learning model and calculates the degree of matching (score). As a result, the machine learning unit 115 can determine whether or not the detection is correct.
  • the lightweight learning data creation unit 116 cuts out a partial image to be used for learning by the machine learning unit 115 from the image of the training data, and creates lightweight learning data including the partial image. Specifically, the lightweight learning data creation unit 116 acquires learning data from the learning data storage unit 121. The lightweight learning data creation unit 116 extracts a partial image including the object 2 from the image of the training data using information such as the position / orientation and size of the object included in the label, associates the partial image with the label, and associates the partial image with the label. Create lightweight training data including partial images and labels. The labels included in the lightweight training data do not have to include the information used to crop the partial image.
  • Multiple partial images may be cut out from one image.
  • no partial image may be cut out from one image. This is when the image was stored in the training dataset but no object was found on the image, or the object was found but the user chooses not to label it. Because there is.
  • FIG. 5 is a diagram showing an example of extracting a partial image.
  • the lightweight learning data creation unit 116 extracts the partial image G2 from the image G1. Then, the lightweight learning data creating unit 116 associates the extracted partial image with the label given by the labeling unit 113, and uses the partial image and the label as learning data.
  • the lightweight learning data creation unit 116 may further perform image processing on the extracted partial image and then store it as lightweight learning data. For example, when the lightweight learning data creation unit 116 inputs the data obtained by reducing the partial image, the data obtained by extracting the features from the partial image, or the like to the machine learning unit 115, the data subjected to such image processing is lightweight. Store as training data. As a result, the lightweight learning data creation unit 116 can reduce the size of the data and reduce the amount of calculation during learning.
  • the lightweight learning data creation unit 116 uses a set of lightweight learning data including partial images and labels as a lightweight learning data set.
  • the machine learning unit 115 inputs each pixel value of the partial image into the learning model and calculates the degree of matching (score).
  • the degree of coincidence is a value from 0 to 1.
  • the machine learning unit 115 calculates the error from the calculated degree of agreement (score), assuming that the label of the detection result is 1.0 if the answer is correct and 0.0 if the label is incorrect.
  • the machine learning unit 115 backpropagates the error in the learning model and updates the parameters (for example, weights) of the learning model. Then, the machine learning unit 115 repeats such a process as many times as the number of detection results (N) used for learning.
  • the learning data control unit 117 associates the lightweight learning data created by the lightweight learning data creation unit 116 with the image processing program and stores it in the storage unit 12. Specifically, the learning data control unit 117 stores the lightweight learning data in a file constituting the image processing program. Since the lightweight learning data has a small file size, it can be stored in a file constituting an image processing program.
  • the learning data control unit 117 may store the lightweight learning data as one or more files in the lightweight learning data storage unit 122, and store the file path to the lightweight learning data in the file of the image processing program.
  • the image processing unit 114 performs image processing on the image data using an image processing program.
  • the layer processing program is stored in the storage unit 12.
  • the image processing program is a program for executing image processing desired by the user.
  • the image processing program may detect the object 2 using the model pattern and determine whether or not the detected region is correct detection.
  • the processing by such an image processing program is disclosed in, for example, Japanese Patent Application Laid-Open No. 2018-151843 (Patent Document 2).
  • the learning data control unit 117 stores the lightweight learning data in association with the image processing program, so that the lightweight learning data can be stored when the learning model stored in the image processing program is re-learned or when additional learning is performed. You can use it to learn.
  • the machine learning unit 115 performs additional learning, it learns using a new learning data set and an existing lightweight learning data set.
  • the lightweight training data may be stored in association with the training model.
  • the learning data control unit 117 deletes the learning data from the learning data storage unit 121 after the lightweight learning data creation unit 116 creates the lightweight learning data.
  • the machine learning device 10 can reduce the size of the storage area required for machine learning.
  • the learning data control unit 117 may delete an unlabeled image from the training data after the lightweight learning data creation unit 116 creates the lightweight learning data.
  • the learning data control unit 117 may select the learning data in the learning data set and delete the selected learning data when the remaining storage area of the storage unit 12 is small. Any method can be used to select the training data. For example, the learning data control unit 117 may delete old learning data. Even if the training data is deleted, the lightweight training data remains, so that the lightweight training data can be retrained. Further, the lightweight learning data creation unit 116 may create lightweight learning data when the learning data control unit 117 deletes the learning data.
  • the display control unit 118 displays the lightweight learning data on the display device 40.
  • the display of the partial image by the display control unit 118 can be performed by a known method (see, for example, Japanese Patent Application Laid-Open No. 2017-151813).
  • the display control unit 118 can display a partial image of the learning data set including the lightweight learning data at high speed.
  • FIG. 6 is a flowchart showing a flow of processing using lightweight learning data in the machine learning device 10.
  • the lightweight learning data creation unit 116 acquires learning data from the learning data storage unit 121.
  • step S2 the lightweight learning data creation unit 116 extracts a partial image including the object 2 from the image of the training data.
  • the lightweight learning data creation unit 116 associates the partial image and the label to create lightweight learning data including the partial image and the label.
  • step S4 the lightweight learning data creation unit 116 determines whether or not lightweight learning data has been created from all the learning data in the learning data storage unit 121.
  • the process proceeds to step S5.
  • the lightweight learning data is not created from all the training data (NO)
  • the process proceeds to step S6.
  • step S5 the machine learning unit 115 performs machine learning using the lightweight learning data.
  • step S6 the learning data control unit 117 stores the lightweight learning data created by the lightweight learning data creation unit 116 in the storage unit 12 in association with the image processing program.
  • the machine learning device 10 performs image processing on the image by using the machine learning unit 115 for learning the learning data including the image and the label for the image and the image processing program for the image.
  • the machine learning unit 115 includes a learning data control unit 117 that is associated and stored, and the machine learning unit 115 learns learning data or lightweight learning data.
  • the machine learning device 10 does not need to cut out a partial image by using the lightweight learning data when learning again, and the learning can be performed at high speed. Further, the machine learning device 10 can reduce the size of the learning data set stored in the storage unit 12 by storing the lightweight learning data. Therefore, the machine learning device 10 can reduce the weight of the learning data and perform learning at high speed.
  • the learning data control unit 117 deletes the learning data after the lightweight learning data creation unit 116 creates the lightweight learning data. As a result, the machine learning device 10 can reduce the size of the storage area required for machine learning. Further, the learning data control unit 117 deletes an unlabeled image from the training data after the lightweight learning data creation unit 116 creates the lightweight learning data. As a result, the machine learning device 10 can reduce the size of the storage area required for machine learning.
  • the machine learning device 10 further includes a display control unit 118 that displays lightweight learning data on the display device 40.
  • the machine learning device 10 can display a partial image of the learning data set including the lightweight learning data at high speed.
  • the learning data control unit 117 stores the lightweight learning data in a file constituting the image processing program. Since the file size of the lightweight learning data is small, it can be stored in the image processing program. As a result, the machine learning device 10 can handle the learning data set in the same file as the image processing program, and can improve the convenience of the user.
  • the learning data control unit 117 may store the lightweight learning data as one or more files in the lightweight learning data storage unit 122, and store the file path to the lightweight learning data in the file of the image processing program.
  • the machine learning device 10 can handle a lightweight data set and can improve the convenience of the user.
  • one machine learning device 10 may be a machine learning system in which a plurality of machine learning devices 10 exist.
  • the learning model stored by any of the machine learning devices 10 may be shared with other machine learning devices 10. If the learning model is shared by a plurality of machine learning devices 10, each machine learning device 10 can perform learning in a distributed manner, so that the machine learning system can improve the learning efficiency. It becomes.
  • the learning data set including the lightweight learning data stored in any of the machine learning devices 10 is shared with the other machine learning devices 10. May be good.
  • the machine learning system can reduce the load on the network by sharing lightweight learning data instead of learning data.
  • the learning data control unit 117 stores the lightweight learning data in association with the image processing program, but the learning data control unit 117 has a learning model for learning the learning data and is lightweight. It may be stored in association with a lightweight learning model for learning training data.
  • the above machine learning device can be realized by hardware, software, or a combination thereof. Further, the control method performed by the above-mentioned machine learning device can also be realized by hardware, software, or a combination thereof.
  • what is realized by software means that it is realized by a computer reading and executing a program.
  • Non-transient computer-readable media include various types of tangible storage media (tangible studio media).
  • Examples of non-temporary computer-readable media include magnetic recording media (eg, hard disk drives), optomagnetic recording media (eg, optomagnetic disks), CD-ROMs (Read Only Memory), CD-Rs, CD-Rs /. W, including semiconductor memory (for example, mask ROM, PROM (Programmable ROM), EPROM (Erasable PROM), flash ROM, RAM (random access memory)).

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Software Systems (AREA)
  • Quality & Reliability (AREA)
  • Computing Systems (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • Medical Informatics (AREA)
  • Databases & Information Systems (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Multimedia (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Image Analysis (AREA)
PCT/JP2021/018191 2020-05-18 2021-05-13 機械学習装置及び機械学習システム Ceased WO2021235311A1 (ja)

Priority Applications (4)

Application Number Priority Date Filing Date Title
CN202180035638.7A CN115668283A (zh) 2020-05-18 2021-05-13 机器学习装置和机器学习系统
JP2022524417A JPWO2021235311A1 (https=) 2020-05-18 2021-05-13
US17/998,351 US20230186457A1 (en) 2020-05-18 2021-05-13 Machine-learning device and machine-learning system
DE112021002846.4T DE112021002846T5 (de) 2020-05-18 2021-05-13 Maschinenlernvorrichtung und Maschinenlernsystem

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2020-086733 2020-05-18
JP2020086733 2020-05-18

Publications (1)

Publication Number Publication Date
WO2021235311A1 true WO2021235311A1 (ja) 2021-11-25

Family

ID=78708353

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2021/018191 Ceased WO2021235311A1 (ja) 2020-05-18 2021-05-13 機械学習装置及び機械学習システム

Country Status (5)

Country Link
US (1) US20230186457A1 (https=)
JP (1) JPWO2021235311A1 (https=)
CN (1) CN115668283A (https=)
DE (1) DE112021002846T5 (https=)
WO (1) WO2021235311A1 (https=)

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2018132962A (ja) * 2017-02-15 2018-08-23 オムロン株式会社 画像出力装置及び画像出力方法
WO2018173800A1 (ja) * 2017-03-21 2018-09-27 日本電気株式会社 画像処理装置、画像処理方法及び記録媒体

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP6348093B2 (ja) 2015-11-06 2018-06-27 ファナック株式会社 入力データから検出対象物の像を検出する画像処理装置および方法
JP6333871B2 (ja) 2016-02-25 2018-05-30 ファナック株式会社 入力画像から検出した対象物を表示する画像処理装置
JP6542824B2 (ja) 2017-03-13 2019-07-10 ファナック株式会社 入力画像から検出した対象物の像の尤度を計算する画像処理装置および画像処理方法
US10632023B2 (en) * 2017-06-13 2020-04-28 The Procter & Gamble Company Systems and methods for inspecting absorbent articles on a converting line
JP6705777B2 (ja) * 2017-07-10 2020-06-03 ファナック株式会社 機械学習装置、検査装置及び機械学習方法
US11030486B2 (en) * 2018-04-20 2021-06-08 XNOR.ai, Inc. Image classification through label progression
JP7129669B2 (ja) * 2018-07-20 2022-09-02 株式会社エヌテック ラベル付き画像データ作成方法、検査方法、プログラム、ラベル付き画像データ作成装置及び検査装置

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2018132962A (ja) * 2017-02-15 2018-08-23 オムロン株式会社 画像出力装置及び画像出力方法
WO2018173800A1 (ja) * 2017-03-21 2018-09-27 日本電気株式会社 画像処理装置、画像処理方法及び記録媒体

Also Published As

Publication number Publication date
JPWO2021235311A1 (https=) 2021-11-25
CN115668283A (zh) 2023-01-31
DE112021002846T5 (de) 2023-03-02
US20230186457A1 (en) 2023-06-15

Similar Documents

Publication Publication Date Title
US11400598B2 (en) Information processing apparatus, method, and robot system
JP7481427B2 (ja) 取り出しシステム及び方法
CN114599488B (zh) 机器学习数据生成装置、机器学习装置、作业系统、计算机程序、机器学习数据生成方法及作业机的制造方法
RU2700246C1 (ru) Способ и система захвата объекта с помощью роботизированного устройства
US10964057B2 (en) Information processing apparatus, method for controlling information processing apparatus, and storage medium
JP6795093B2 (ja) 判定装置、判定方法及び判定プログラム
JP7377627B2 (ja) 物体検出装置、物体把持システム、物体検出方法及び物体検出プログラム
CN114080590B (zh) 使用先进扫描技术的机器人料箱拾取系统和方法
JP2018036770A (ja) 位置姿勢推定装置、位置姿勢推定方法、及び位置姿勢推定プログラム
JP2020077231A (ja) 位置検出プログラム、位置検出方法及び位置検出装置
US20160110840A1 (en) Image processing method, image processing device, and robot system
CN117260702A (zh) 用于控制机器人来操纵、尤其是拾取对象的方法
JP6237122B2 (ja) ロボット、画像処理方法及びロボットシステム
JP5083715B2 (ja) 三次元位置姿勢計測方法および装置
WO2021235311A1 (ja) 機械学習装置及び機械学習システム
CN114505864A (zh) 一种手眼标定方法、装置、设备及存储介质
US12459130B2 (en) Robot operation system, robot operation method, and program
CN116137062B (zh) 生成训练数据的方法、系统以及存储介质
JP7376446B2 (ja) 作業分析プログラム、および、作業分析装置
KR20240096990A (ko) 비고정 물체를 위치 이동시키는 로봇의 제어 장치
JP2013053920A (ja) ３次元物体位置検出装置、そのプログラム
CN118691524A (zh) 工业3d视觉识别方法、装置、设备及存储介质
JP7420917B2 (ja) 機械学習装置
US10379620B2 (en) Finger model verification method and information processing apparatus
US20250123678A1 (en) Cross-reality device, storage medium, processing device, generation method, and processing method

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21808820

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2022524417

Country of ref document: JP

Kind code of ref document: A

122 Ep: pct application non-entry in european phase

Ref document number: 21808820

Country of ref document: EP

Kind code of ref document: A1