JP7560393B2 - Distributedness determination method, distributedness determination device, and trained model generation method - Google Patents
Distributedness determination method, distributedness determination device, and trained model generation method Download PDFInfo
- Publication number
- JP7560393B2 JP7560393B2 JP2021043856A JP2021043856A JP7560393B2 JP 7560393 B2 JP7560393 B2 JP 7560393B2 JP 2021043856 A JP2021043856 A JP 2021043856A JP 2021043856 A JP2021043856 A JP 2021043856A JP 7560393 B2 JP7560393 B2 JP 7560393B2
- Authority
- JP
- Japan
- Prior art keywords
- dispersibility
- image
- flow curve
- particle dispersion
- determination
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims description 52
- 239000002245 particle Substances 0.000 claims description 123
- 239000006185 dispersion Substances 0.000 claims description 62
- 238000001595 flow curve Methods 0.000 claims description 59
- 238000012549 training Methods 0.000 claims description 21
- 238000000605 extraction Methods 0.000 claims description 16
- 239000007788 liquid Substances 0.000 claims description 16
- 238000005259 measurement Methods 0.000 claims description 9
- 238000003384 imaging method Methods 0.000 description 18
- 238000012545 processing Methods 0.000 description 9
- 238000010801 machine learning Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 239000002612 dispersion medium Substances 0.000 description 4
- 239000000284 extract Substances 0.000 description 4
- 239000000463 material Substances 0.000 description 4
- ZWEHNKRNPOVVGH-UHFFFAOYSA-N 2-Butanone Chemical compound CCC(C)=O ZWEHNKRNPOVVGH-UHFFFAOYSA-N 0.000 description 3
- 238000013135 deep learning Methods 0.000 description 3
- 239000012530 fluid Substances 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 239000011236 particulate material Substances 0.000 description 3
- 239000011248 coating agent Substances 0.000 description 2
- 238000000576 coating method Methods 0.000 description 2
- 239000011521 glass Substances 0.000 description 2
- 239000004925 Acrylic resin Substances 0.000 description 1
- 229920000178 Acrylic resin Polymers 0.000 description 1
- LFQSCWFLJHTTHZ-UHFFFAOYSA-N Ethanol Chemical compound CCO LFQSCWFLJHTTHZ-UHFFFAOYSA-N 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000007865 diluting Methods 0.000 description 1
- 150000004679 hydroxides Chemical class 0.000 description 1
- 239000010954 inorganic particle Substances 0.000 description 1
- 238000000691 measurement method Methods 0.000 description 1
- 239000003094 microcapsule Substances 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000011146 organic particle Substances 0.000 description 1
- 239000003960 organic solvent Substances 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 239000011347 resin Substances 0.000 description 1
- 229920005989 resin Polymers 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 230000009974 thixotropic effect Effects 0.000 description 1
- -1 type Substances 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
Landscapes
- Image Analysis (AREA)
Description
本発明は、分散性判定方法、分散性判定装置および学習済みモデルの生成方法に関する。 The present invention relates to a method for determining variance, a device for determining variance, and a method for generating a trained model.
粒子分散液の分散性を判定する分散性判定方法が知られている（例えば、特許文献１参照）。 A method for determining the dispersibility of a particle dispersion is known (see, for example, Patent Document 1).
特許文献１に記載された測定方法（分散性判定方法）では、粒子分散材料（粒子分散液）内部の粒子状材料の凝集構造を異方性信号として測定しなければならないため、工程が煩雑になるという不都合がある。 The measurement method (dispersibility determination method) described in Patent Document 1 has the disadvantage that the process is complicated because the aggregate structure of the particulate material inside the particle dispersion material (particle dispersion liquid) must be measured as an anisotropic signal.
本発明の目的は、簡易な工程で分散性を判定することができる分散性判定方法、分散性判定装置および学習済みモデルの生成方法を提供することにある。 The object of the present invention is to provide a method for determining dispersibility, a device for determining dispersibility, and a method for generating a trained model that can determine dispersibility through simple steps.
本発明の一態様に係る分散性判定方法は、粒子分散液の分散性を判定する分散性判定方法であって、前記粒子分散液を撮像した撮像画像を取得する画像取得工程と、学習用の粒子分散液を撮像した教師画像と、前記学習用の粒子分散液について測定したフローカーブの形状特性から判定される前記分散性とを教師データとして機械学習させた学習済みモデルを用いて、前記撮像画像から前記分散性を判定する分散性判定工程とを実施する。 A dispersibility determination method according to one aspect of the present invention is a dispersibility determination method for determining the dispersibility of a particle dispersion, which includes an image acquisition step of acquiring an image of the particle dispersion, and a dispersibility determination step of determining the dispersibility from the image by using a trained model that has been machine-learned using a teacher image of a training particle dispersion and the dispersibility determined from the shape characteristics of a flow curve measured for the training particle dispersion as training data.
本発明の一態様に係る分散性判定方法において、前記学習済みモデルは、前記撮像画像における粒子の数、最大粒子サイズ、および粒子の大きさの偏差を特徴量として機械学習され、前記撮像画像における粒子の数、最大粒子サイズ、および粒子の大きさの偏差を特徴量として抽出して前記分散性判定工程に引き渡す特徴量抽出工程を実施することが好ましい。 In a dispersibility determination method according to one aspect of the present invention, the trained model is machine-trained using the number of particles, maximum particle size, and particle size deviation in the captured image as feature quantities, and it is preferable to carry out a feature quantity extraction process in which the number of particles, maximum particle size, and particle size deviation in the captured image are extracted as feature quantities and passed to the dispersibility determination process.
本発明の一態様に係る分散性判定方法において、前記フローカーブの形状特性は、前記フローカーブにおけるせん断速度が０．１ｓ－１のときの粘度と、せん断速度が１０ｓ－１のときの粘度との比であることが好ましい。 In the dispersibility determination method according to one aspect of the present invention, the shape characteristic of the flow curve is preferably a ratio of the viscosity when the shear rate in the flow curve is 0.1 s −1 to the viscosity when the shear rate is 10 s −1 .
本発明の一態様に係る分散性判定装置は、粒子分散液の分散性を判定する分散性判定装置であって、前記粒子分散液を撮像した撮像画像を取得する画像取得手段と、学習用の粒子分散液を撮像した教師画像と、前記学習用の粒子分散液について測定したフローカーブの形状特性から判定される前記分散性とを教師データとして機械学習させた学習済みモデルを用いて、前記撮像画像から前記分散性を判定する分散性判定手段とを備えている。 A dispersibility determination device according to one aspect of the present invention is a dispersibility determination device that determines the dispersibility of a particle dispersion, and includes an image acquisition means for acquiring an image of the particle dispersion, and a dispersibility determination means for determining the dispersibility from the image by using a trained model that has been machine-learned using a teacher image of a learning particle dispersion and the dispersibility determined from the shape characteristics of a flow curve measured for the learning particle dispersion as teacher data.
本発明の一態様に係る学習済みモデルの生成方法は、粒子分散液の分散性を判定するための学習済みモデルの生成方法であって、学習用の粒子分散液のフローカーブを測定するフローカーブ測定工程と、前記学習用の粒子分散液を撮像して教師画像とする教師画像撮像工程と、前記フローカーブ測定工程で測定された前記フローカーブの形状特性から前記分散性を判定するフローカーブ判定工程と、記教師画像撮像工程で撮像された前記教師画像と、前記フローカーブ判定工程で判定された前記分散性とを教師データとして機械学習させ、前記学習済みモデルを生成するモデル生成工程とを実施する。 A trained model generation method according to one aspect of the present invention is a trained model generation method for determining the dispersibility of a particle dispersion, which includes a flow curve measurement step of measuring the flow curve of the training particle dispersion, a teacher image capture step of capturing an image of the training particle dispersion to create a teacher image, a flow curve determination step of determining the dispersibility from the shape characteristics of the flow curve measured in the flow curve measurement step, and a model generation step of machine learning the teacher image captured in the teacher image capture step and the dispersibility determined in the flow curve determination step as teacher data to generate the trained model.
本発明の一態様に係る学習済みモデルの生成方法において、前記教師画像における粒子の数、最大粒子サイズ、および粒子の大きさの偏差を特徴量として抽出する特徴量抽出工程を実施し、前記モデル生成工程では、前記特徴量抽出工程で抽出された粒子の数、最大粒子サイズ、および粒子の大きさの偏差を特徴量として機械学習させることが好ましい。 In a method for generating a trained model according to one aspect of the present invention, it is preferable to carry out a feature extraction step of extracting the number of particles, maximum particle size, and particle size deviation in the teacher image as features, and in the model generation step, machine learning is performed on the number of particles, maximum particle size, and particle size deviation extracted in the feature extraction step as features.
本発明の一態様に係る学習済みモデルの生成方法において、前記フローカーブの形状特性は、前記フローカーブにおけるせん断速度が０．１ｓ－１のときの粘度と、せん断速度が１０ｓ－１のときの粘度との比であることが好ましい。 In the method for generating a trained model according to one embodiment of the present invention, the shape characteristic of the flow curve is preferably a ratio of the viscosity when the shear rate in the flow curve is 0.1 s -1 to the viscosity when the shear rate is 10 s -1 .
本発明の一態様によれば、粒子分散液を撮像すれば、その撮像画像から分散性を判定することができるので、粒子状材料の凝集構造を異方性信号として測定しなくてもよく、簡易な工程で分散性を判定することができる。 According to one aspect of the present invention, by capturing an image of a particle dispersion, the dispersibility can be determined from the captured image, so there is no need to measure the aggregate structure of the particulate material as an anisotropic signal, and dispersibility can be determined through a simple process.
以下、本発明の一実施形態を図面に基づいて説明する。
［装置構成］
図１において、分散性判定装置ＥＡは、粒子分散液の分散性を判定する装置であって、パーソナルコンピュータやサーバ等のコンピュータにより構成されている。なお、粒子分散液は、分散媒である液体に固体または液体の分散質が粒子となって分散したものである。
分散性判定装置ＥＡは、記憶手段１０と、処理手段２０と、操作手段３０とを備え、撮像手段４０と、出力手段５０とで分散性判定システムＥＡ１を構成している。
Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
[Device configuration]
1, the dispersibility determination device EA is a device for determining the dispersibility of a particle dispersion liquid, and is configured with a computer such as a personal computer, a server, etc. Note that a particle dispersion liquid is a liquid dispersion medium in which solid or liquid dispersoids are dispersed as particles.
The dispersibility determination device EA includes a storage means 10, a processing means 20, and an operation means 30, and together with an imaging means 40 and an output means 50 constitute a dispersibility determination system EA1.
記憶手段１０は、メモリやハードディスク等により構成され、分散性判定装置ＥＡおよび分散性判定システムＥＡ１を制御する各種プログラムを記憶している。また、記憶手段１０は、粒子分散液の分散性を判定するための学習済みモデルを記憶している。 The storage means 10 is composed of a memory, a hard disk, etc., and stores various programs that control the dispersibility determination device EA and the dispersibility determination system EA1. The storage means 10 also stores a trained model for determining the dispersibility of a particle dispersion.
処理手段２０は、ＣＰＵ（Central Processing Unit）やＧＰＵ（Central Graphics Processing Unit）等のプロセッサにより構成され、画像取得手段２１と、特徴量抽出手段２２と、分散性判定手段２３とを備えている。
画像取得手段２１は、粒子分散液を撮像した撮像画像を撮像手段４０から取得する。
特徴量抽出手段２２は、画像取得手段２１で取得された撮像画像から、所定の特徴量を抽出して分散性判定手段２３に引き渡す。特徴量抽出手段２２が抽出する特徴量としては、粒子分散液の撮像画像中の粒子の数、最大粒子サイズ、粒子の大きさの偏差、粒子の最大周囲長、粒子の大きさの平均、撮像画像を分割した際の各分割領域における粒子の大きさの個数分布、粒子の周囲長の平均、粒子の周囲長の偏差等が例示できる。
分散性判定手段２３は、記憶手段１０に記憶された学習済みモデルを用いて、画像取得手段２１で取得された撮像画像から粒子分散液の分散性を判定する。
The
The image acquisition means 21 acquires an image of the particle dispersion liquid from the imaging means 40 .
The feature extraction means 22 extracts predetermined feature amounts from the captured image acquired by the image acquisition means 21, and passes them to the dispersibility determination means 23. Examples of feature amounts extracted by the feature extraction means 22 include the number of particles in the captured image of the particle dispersion, the maximum particle size, the deviation of particle size, the maximum perimeter of particles, the average particle size, the number distribution of particle sizes in each divided region when the captured image is divided, the average perimeter of particles, the deviation of particle perimeter, and the like.
The dispersibility determination means 23 uses the trained model stored in the storage means 10 to determine the dispersibility of the particle dispersion from the captured image acquired by the image acquisition means 21 .
操作手段３０は、キーボード、操作パネル、タッチパネル、マウス、各種スイッチ、音声入力操作用のマイク等により構成され、各種操作による操作信号を分散性判定装置ＥＡに入力可能とされている。 The operation means 30 is composed of a keyboard, an operation panel, a touch panel, a mouse, various switches, a microphone for voice input operation, etc., and can input operation signals from various operations to the dispersibility determination device EA.
撮像手段４０は、カメラ、撮影機、撮像機能付き顕微鏡、イメージングセンサ等により構成され、撮像画像を処理手段２０に送信可能となっている。 The imaging means 40 is composed of a camera, a camera, a microscope with imaging function, an imaging sensor, etc., and is capable of transmitting the captured image to the processing means 20.
出力手段５０は、ディスプレイやパネル等の表示装置や、表示灯やスピーカー等の報知装置等により構成され、分散性判定装置ＥＡによる判定結果を画面に出力したり、点灯や音等で出力したりするようになっている。 The output means 50 is composed of a display device such as a display or panel, or an alarm device such as an indicator light or speaker, and is configured to output the judgment result by the dispersibility judgment device EA on a screen or by lighting up or sound, etc.
［学習済みモデル］
学習済みモデルは、学習用の粒子分散液を撮像した教師画像と、学習用の粒子分散液について測定したフローカーブの形状特性から判定される分散性とを教師データとして機械学習させたものであり、図２に示す以下の手順で生成される。なお、フローカーブとは、粒子分散液のせん断速度と粘度との関係、または粒子分散液のせん断速度と応力との関係をグラフ化したものである。
[Trained model]
The trained model is machine-learned using a teacher image of a training particle dispersion and dispersibility determined from the shape characteristics of a flow curve measured for the training particle dispersion as training data, and is generated by the following procedure shown in Fig. 2. Note that a flow curve is a graph of the relationship between the shear rate and viscosity of a particle dispersion, or the relationship between the shear rate and stress of a particle dispersion.
先ず、学習用に様々な種類、状態の粒子分散液を用意し、これらの学習用の粒子分散液についてフローカーブを測定する（ステップＳＴ１１）。この際使用する測定機器としては、共軸二重円筒型粘度計、Ｂ型粘度計、Ｅ型粘度計（コーンプレート型粘度計）等、特に限定されないが、回転式粘度計であれば、回転数を自在に変化させながら粘度を測定することができるレオメータが好ましい。 First, various types and states of particle dispersions are prepared for learning, and the flow curves of these particle dispersions for learning are measured (step ST11). The measuring equipment used in this case is not particularly limited, and may be a coaxial double cylinder viscometer, a B-type viscometer, an E-type viscometer (cone-plate type viscometer), etc., but if it is a rotational viscometer, a rheometer that can measure viscosity while freely changing the rotation speed is preferable.
次いで、学習用の粒子分散液を撮像して教師画像とする（ステップＳＴ１２）。撮像は、撮像手段４０で行ってもよいし、撮像手段４０以外のもので行ってもよく、例えば、２枚のスライドガラスで挟み込まれて塗膜となった粒子分散液を、撮像機能を有するデジタル顕微鏡で撮像してもよい。 Next, the particle dispersion for learning is imaged and used as a teacher image (step ST12). The image may be captured by the imaging means 40 or by something other than the imaging means 40. For example, the particle dispersion that has been sandwiched between two glass slides and turned into a coating film may be imaged by a digital microscope with an imaging function.
その後、ステップＳＴ１１で測定されたフローカーブの形状特性から、学習用の粒子分散液の分散性を判定する（ステップＳＴ１３）。このように判断する理由として、粒子分散液は、完全分散状態に近ければ、図３で符号ＦＣ１を付したフローカーブが示すように、低せん断領域から高せん断領域にわたって粘度があまり変化しないニュートン流体となっており、その特徴がフローカーブの形状特性に表れるからである。分散性の判定に用いるフローカーブの形状特性としては、例えば、フローカーブの任意の点における傾きまたはこれに相当する特性が例示できる。
本実施形態の場合、フローカーブにおけるせん断速度が０．１ｓ－１のときの粘度と、せん断速度が１０ｓ－１のときの粘度との比をフローカーブの形状特性として、分散性を判断する。具体的には、図３で符号ＦＣ１を付したフローカーブのように、下記式（１）で求まる指標ＲＩの値がＲＩ≧０．９であれば、粒子分散液がニュートン流体であるとして分散性良（合格）と判断する。一方、図３で符号ＦＣ２を付したフローカーブのように、ＲＩ＜０．９であれば、粒子分散液がチキソトロピー流体であるとして分散性否（不合格）と判断する。
Thereafter, the dispersibility of the learning particle dispersion is judged from the shape characteristics of the flow curve measured in step ST11 (step ST13). The reason for this judgment is that if the particle dispersion is close to a completely dispersed state, it becomes a Newtonian fluid whose viscosity does not change much from the low shear region to the high shear region, as shown by the flow curve marked with the symbol FC1 in Fig. 3, and this characteristic is reflected in the shape characteristics of the flow curve. The shape characteristics of the flow curve used to judge the dispersibility can be, for example, the slope at any point of the flow curve or a characteristic equivalent thereto.
In this embodiment, the dispersibility is judged by the ratio of the viscosity at a shear rate of 0.1 s -1 to the viscosity at a shear rate of 10 s -1 in the flow curve, which is taken as the shape characteristic of the flow curve. Specifically, as in the flow curve marked with reference FC1 in Fig. 3, if the index RI value calculated by the following formula (1) is RI≧0.9, the particle dispersion is deemed to be a Newtonian fluid and is judged to have good dispersibility (pass). On the other hand, as in the flow curve marked with reference FC2 in Fig. 3, if RI<0.9, the particle dispersion is deemed to be a thixotropic fluid and is judged to have poor dispersibility (fail).
［式１］
ＲＩ＝せん断速度が１０ｓ－１のときの粘度／せん断速度が０．１ｓ－１のときの粘度・・・（１）
[Formula 1]
RI = viscosity at a shear rate of 10 s -1 / viscosity at a shear rate of 0.1 s -1 ... (1)
次に、ステップＳＴ１２で撮像した教師画像から、所定の特徴量を抽出する（ステップＳＴ１４）。本実施形態の場合、少なくとも粒子の数、最大粒子サイズ、および粒子の大きさの偏差を特徴量として抽出する。なお、特徴量を抽出する際は、各撮像画像に対して、輝度値の平均、分散に基づく閾値を設定し、二値化処理を施しておくとよい。 Next, predetermined feature amounts are extracted from the teacher image captured in step ST12 (step ST14). In this embodiment, at least the number of particles, the maximum particle size, and the deviation in particle size are extracted as feature amounts. When extracting feature amounts, it is advisable to set a threshold based on the average and variance of the brightness values for each captured image and perform binarization processing.
そして、教師画像から抽出した特徴量と、ステップＳＴ１３で判定された分散性とを教師データとして機械学習させ、粒子分散液の撮像画像を入力、粒子分散液の分散性を出力とする学習済みモデルを生成する（ステップＳＴ１５）。機械学習の具体的な手法としては、公知の各種の技術が利用可能であり、特に限定されない。本実施形態の場合、機械学習のアルゴリズムに、ニューラルネットワーク、サポートベクターマシン、最近傍法（ニアレストネイバー法）、ランダムフォレスト、確率的勾配法、およびカーネル近似を使用する。 Then, machine learning is performed using the feature amount extracted from the training image and the dispersibility determined in step ST13 as training data, and a trained model is generated in which the captured image of the particle dispersion is input and the dispersibility of the particle dispersion is output (step ST15). Specific machine learning techniques include various well-known techniques and are not particularly limited. In this embodiment, the machine learning algorithms used are neural networks, support vector machines, nearest neighbor methods, random forests, stochastic gradient methods, and kernel approximations.
［分散性判定方法］
以上の分散性判定装置ＥＡを備えた分散性判定システムＥＡ１の場合を例として、図４に示す以下の手順で実施される分散性判定方法を説明する。
先ず、分散性判定システムＥＡ１に対し、当該分散性判定システムＥＡ１の使用者（以下、単に「使用者」という）が、操作手段３０を介して自動運転開始の信号を入力する。次いで、使用者または、多関節ロボットやベルトコンベア等の図示しない搬送手段が、例えば、２枚のスライドガラスで挟み込まれて塗膜となった粒子分散液や、透明な容器に収容された粒子分散液を所定の位置に配置すると、処理手段２０が撮像手段４０を駆動し、粒子分散液を撮像する（ステップＳＴ２１）。その後、画像取得手段２１は、撮像手段４０から粒子分散液の撮像画像を取得する（ステップＳＴ２２）。
[Method for determining dispersibility]
A dispersibility determination method performed in the following procedure shown in FIG. 4 will be described using a dispersibility determination system EA1 equipped with the above-described dispersibility determination device EA as an example.
First, a user of the dispersibility determination system EA1 (hereinafter, simply referred to as "user") inputs a signal to start automatic operation to the dispersibility determination system EA1 via the operation means 30. Next, when the user or a conveying means (not shown), such as an articulated robot or a belt conveyor, places a particle dispersion, for example, a coating film sandwiched between two slide glasses or a particle dispersion contained in a transparent container, at a predetermined position, the processing means 20 drives the imaging means 40 to capture an image of the particle dispersion (step ST21). After that, the image acquisition means 21 acquires an image of the particle dispersion from the imaging means 40 (step ST22).
次に、特徴量抽出手段２２は、ステップＳＴ２２で取得された撮像画像から所定の特徴量を抽出して分散性判定手段２３に引き渡す（ステップＳＴ２３）。この際抽出する特徴量は、学習済みモデルの作成時に教師画像から抽出した特徴量と同じにする。そして、分散性判定手段２３は、学習済みモデルを用いて撮像画像から分散性を判定し、判定結果を出力手段５０に出力する（ステップＳＴ２４）。本実施形態の場合、分散性判定手段２３は、分散性を２つのレベルに分け、分散性良（合格）または分散性否（不合格）の判定を行う。次いで、出力手段５０は、分散性の判定結果を表示装置に表示したり、報知装置を点灯したり、報知装置から音を出したりして使用者に知らせる（ステップＳＴ２５）。 Next, the feature extraction means 22 extracts a predetermined feature from the captured image acquired in step ST22 and passes it to the dispersibility determination means 23 (step ST23). The feature extracted at this time is the same as the feature extracted from the teacher image when the trained model was created. The dispersibility determination means 23 then uses the trained model to determine dispersibility from the captured image and outputs the determination result to the output means 50 (step ST24). In this embodiment, the dispersibility determination means 23 divides dispersibility into two levels and determines whether the dispersibility is good (pass) or not (fail). Next, the output means 50 notifies the user of the dispersibility determination result by displaying it on a display device, turning on an alarm device, or emitting a sound from the alarm device (step ST25).
［判定例］
アクリル系樹脂をメチルエチルケトンで希釈した溶液に、表１に示す分散質を添加、分散した粒子分散液のサンプルＡ～Ｈの教師画像を用いて学習済みモデルを作成し、当該学習済みモデルを用いて、表１のサンプルＪ～Ｍの分散性判定を行った。
[Judgment example]
The dispersoids shown in Table 1 were added to a solution obtained by diluting an acrylic resin with methyl ethyl ketone, and a trained model was created using teacher images of samples A to H of the dispersed particle dispersion liquid. The trained model was then used to judge the dispersibility of samples J to M in Table 1.
先ず、サンプルＡ～Ｈのフローカーブ測定を行った結果、各サンプルの所定時間経過後の指標ＲＩの値は、下記表２に示す値となった。 First, flow curve measurements were performed on samples A to H, and the RI index values for each sample after a specified time had elapsed were as shown in Table 2 below.
次いで、教師画像を得るために、サンプルＡ～Ｈをデジタル顕微鏡で撮像した後、各教師画像に二値化処理を施した。その後、各教師画像における粒子の数、最大粒子サイズ、粒子の大きさの偏差、粒子の最大周囲長、粒子の大きさの平均、撮像画像を分割した際の各分割領域における粒子の大きさの個数分布、粒子の周囲長の平均、粒子の周囲長の偏差を特徴量として抽出した。次に、表２でＲＩ≧０．９となったサンプルの教師画像に分散性良（合格）の判定結果を、ＲＩ＜０．９のサンプルの教師画像に分散性否（不合格）の判定結果を付した上で、サンプルＡ～Ｈの教師画像および判定結果から学習モデルを生成した。 Next, to obtain teacher images, samples A to H were photographed with a digital microscope, and each teacher image was subjected to a binarization process. The number of particles in each teacher image, maximum particle size, deviation in particle size, maximum perimeter of particles, average particle size, particle size number distribution in each divided region when the photographed image was divided, average perimeter of particles, and deviation in perimeter of particles were then extracted as feature quantities. Next, a judgment result of good dispersibility (pass) was given to the teacher images of samples with RI ≥ 0.9 in Table 2, and a judgment result of poor dispersibility (fail) was given to the teacher images of samples with RI < 0.9, and a learning model was generated from the teacher images and judgment results of samples A to H.
作成された学習済みモデルを用いてサンプルＪ～Ｍの分散性を判定した結果を表３に、サンプルＪ～Ｍの実際のフローカーブ測定結果から得られる指標ＲＩの値を表４に示す。 Table 3 shows the results of judging the dispersibility of samples J to M using the trained model that was created, and Table 4 shows the index RI values obtained from the actual flow curve measurement results of samples J to M.
表３、４からわかるように、サンプルＪ～Ｍの撮像画像から学習済みモデルを用いて判定した分散性は、サンプルＪ～Ｍの実際のフローカーブ測定結果から得られる指標ＲＩの値と対応し、良好な判定結果が得られた。
また、撮像画像中の特徴量の数を減らし、粒子の数、最大粒子サイズ、および粒子の大きさの偏差を特徴量として同様の判定を行ったところ、判定精度がわずかに低下したものの、実用上問題ない良好な判定結果が得られた。
As can be seen from Tables 3 and 4, the dispersibility determined from the captured images of samples J to M using the trained model corresponds to the value of the index RI obtained from the actual flow curve measurement results of samples J to M, and good determination results were obtained.
In addition, when the number of features in the captured image was reduced and a similar judgment was made using the number of particles, the maximum particle size, and the deviation in particle size as the features, the judgment accuracy decreased slightly, but good judgment results that were sufficient for practical purposes were obtained.
以上のような実施形態によれば、学習用の粒子分散液を撮像した教師画像と、学習用の粒子分散液について測定したフローカーブの形状特性から判定される分散性とを教師データとして機械学習させた学習済みモデルを用いて、撮像画像から分散性を判定する。このため、粒子分散液を撮像すれば分散性を判定することができるので、粒子状材料の凝集構造を異方性信号として測定しなくてもよく、簡易な工程で分散性を判定することができる。 According to the above embodiment, dispersibility is determined from the captured image using a trained model that uses a teacher image of the training particle dispersion and dispersibility determined from the shape characteristics of the flow curve measured for the training particle dispersion as training data. Therefore, dispersibility can be determined by simply capturing an image of the particle dispersion, and there is no need to measure the aggregate structure of the particulate material as an anisotropic signal, making it possible to determine dispersibility through a simple process.
また、特徴量抽出手段２２が、撮像画像における粒子の数、最大粒子サイズ、および粒子の大きさの偏差を特徴量として抽出して分散性判定手段２３に引き渡すため、学習済みモデルによる分散性の判定精度を高めることができる。 In addition, the feature extraction means 22 extracts the number of particles in the captured image, the maximum particle size, and the deviation in particle size as features and passes them on to the dispersibility determination means 23, thereby improving the accuracy of dispersibility determination using the trained model.
また、フローカーブにおけるせん断速度が０．１ｓ－１のときの粘度と、せん断速度が１０ｓ－１のときの粘度との比をフローカーブの形状特性として分散性を判定するため、せん断速度比が１：１０となる位置での粘度比を用いるいわゆるＴｉ値に比べ、フローカーブの形状特性をより広範囲で特定することができ、分散性の判定精度を高めることができる。 In addition, since the dispersibility is determined using the ratio of the viscosity when the shear rate in the flow curve is 0.1 s -1 to the viscosity when the shear rate is 10 s -1 as the shape characteristic of the flow curve, the shape characteristic of the flow curve can be specified over a wider range than the so-called Ti value, which uses the viscosity ratio at a position where the shear rate ratio is 1:10, and the accuracy of dispersibility determination can be improved.
以上のように、本発明を実施するための最良の構成、方法等は、前記記載で開示されているが、本発明は、これに限定されるものではない。すなわち、本発明は、主に特定の実施形態に関して特に図示され、かつ説明されているが、本発明の技術的思想および目的の範囲から逸脱することなく、以上述べた実施形態に対し、形状、材質、数量、その他の詳細な構成において、当業者が様々な変形を加えることができるものである。また、上記に開示した形状、材質などを限定した記載は、本発明の理解を容易にするために例示的に記載したものであり、本発明を限定するものではないから、それらの形状、材質などの限定の一部もしくは全部の限定を外した部材の名称での記載は、本発明に含まれる。 As described above, the best configurations and methods for implementing the present invention have been disclosed in the above description, but the present invention is not limited thereto. In other words, the present invention has been mainly illustrated and described with reference to specific embodiments, but those skilled in the art can make various modifications to the above-described embodiments in terms of shape, material, quantity, and other detailed configurations without departing from the scope of the technical idea and purpose of the present invention. Furthermore, the descriptions limiting the shapes, materials, etc. disclosed above are provided as examples to facilitate understanding of the present invention, and do not limit the present invention. Therefore, descriptions of the names of components that remove some or all of the limitations on the shapes, materials, etc. are included in the present invention.
例えば、記憶手段１０は、分散性判定装置ＥＡに内蔵されていてもよいし、分散性判定装置ＥＡに外付けするタイプであってもよい。 For example, the storage means 10 may be built into the dispersibility determination device EA, or may be of a type that is externally attached to the dispersibility determination device EA.
処理手段２０は、図５に示すように、画像取得手段２１で取得した撮像画像から学習済みモデルを生成するモデル生成手段２４を備え、モデル生成手段２４で生成された学習済みモデルを用いて、分散性判定手段２３で撮像画像から分散性を判定してもよい。即ち、
学習用の粒子分散液を撮像手段４０で撮像した教師画像を画像取得手段２１で取得し、特徴量抽出手段２２で教師画像から所定の特徴量を抽出し、モデル生成手段２４で教師画像を用いて学習済みモデルを生成してもよい。この場合、学習用の粒子分散液のフローカーブは、レオメータ等のフローカーブ測定機器から受信してもよいし、操作手段３０から入力されてもよい。また、学習用の粒子分散液の分散性は、そのフローカーブからモデル生成手段２４が上記式（１）を用いて判定してもよいし、判定結果が操作手段３０から入力されてもよい。なお、モデル生成手段２４は、ディープラーニング（深層学習で）学習済みモデルを生成してもよい。
画像取得手段２１は、粒子分散液の撮像画像を撮像手段４０から直接取得してもよいし、撮像手段４０から直接取得しなくてもよく、例えば、撮像手段４０による撮像画像が蓄積されたサーバやデータベース等の画像蓄積手段から撮像画像を取得してもよい。
特徴量抽出手段２２は、粒子の数、最大粒子サイズ、粒子の大きさの偏差、粒子の最大周囲長、粒子の大きさの平均、撮像画像を分割した際の各分割領域における粒子の大きさの個数分布、粒子の周囲長の平均、粒子の周囲長の偏差のうち、任意の複数項目を特徴量として抽出してもよい。
分散性判定手段２３は、指標ＲＩによる分散性良否（分散性合否）の判定閾値を０．９以外の値としてもよいし、指標ＲＩの値に応じて分散性を３つ以上のレベルに分けて判定してもよく、例えば、ＲＩ≧０．９５となった場合は分散性優（合格）、０．９５＞ＲＩ≧０．９となった場合は分散性良（合格）、ＲＩ＜０．９の場合は分散性否（不合格）と判定してもよいし、ＲＩ＝１．０～０．９の範囲を０．０１ごとに１０分割し、分散性を１０段階で判定してもよいし、ＲＩの値をそのまま判定結果として出力してもよい。
分散性判定手段２３は、フローカーブの任意の点における傾きや、フローカーブの任意の複数点を通る直線の傾きをフローカーブの形状特性とし、これらの傾きを指標ＲＩとして分散性を判定してもよい。この場合、上記の場合と同様に、例えば、フローカーブの傾きが閾値以上なら分散性良（合格）、閾値未満なら分散性否（不合格）と判定してもよいし、フローカーブの傾きに応じて分散性を３つ以上のレベルに分けて判定してもよいし、フローカーブの傾きの値をそのまま判定結果として出力してもよい。
As shown in Fig. 5, the processing means 20 may include a model generating means 24 that generates a trained model from the captured image acquired by the
A teacher image of the particle dispersion for learning captured by the imaging means 40 may be acquired by the image acquisition means 21, a predetermined feature amount may be extracted from the teacher image by the feature amount extraction means 22, and a trained model may be generated by the model generation means 24 using the teacher image. In this case, the flow curve of the particle dispersion for learning may be received from a flow curve measuring device such as a rheometer, or may be input from the operation means 30. The dispersibility of the particle dispersion for learning may be determined by the model generation means 24 from the flow curve using the above formula (1), or the determination result may be input from the operation means 30. The model generation means 24 may generate a trained model by deep learning.
The image acquisition means 21 may acquire the image of the particle dispersion liquid directly from the imaging means 40, or it does not have to acquire the image directly from the imaging means 40, and may acquire the image, for example, from an image storage means such as a server or database in which images acquired by the imaging means 40 are stored.
The feature extraction means 22 may extract as features any multiple of the number of particles, maximum particle size, particle size deviation, maximum particle perimeter, average particle size, particle size number distribution in each divided area when the captured image is divided, average particle perimeter, and deviation of particle perimeter.
The dispersibility determination means 23 may set the determination threshold for determining whether dispersibility is good or bad (dispersibility pass/fail) based on the index RI to a value other than 0.9, or may divide the dispersibility into three or more levels depending on the value of the index RI. For example, if RI≧0.95, the dispersibility may be determined to be excellent (pass), if 0.95>RI≧0.9, the dispersibility may be determined to be good (pass), and if RI<0.9, the dispersibility may be determined to be bad (fail), the range of RI=1.0 to 0.9 may be divided into 10 parts in increments of 0.01 and the dispersibility may be determined on a 10-level scale, or the RI value may be output as the determination result as is.
The dispersibility determination means 23 may determine the dispersibility by using the slope at an arbitrary point of the flow curve or the slope of a straight line passing through arbitrary multiple points of the flow curve as the shape characteristic of the flow curve, and by using these slopes as the index RI. In this case, as in the above case, for example, if the slope of the flow curve is equal to or greater than a threshold value, the dispersibility may be determined as good (passed) and if it is less than the threshold value, the dispersibility may be determined as poor (failed), or the dispersibility may be determined by dividing it into three or more levels according to the slope of the flow curve, or the value of the slope of the flow curve may be output as the determination result as it is.
操作手段３０は、分散性判定装置ＥＡから分離可能に構成されていてもよいし、分離不能に構成されていてもよい。
操作手段３０は、分散性判定装置ＥＡに備わっていてもよいし、備わっていなくてもよく、備わっていない場合、分散性判定装置ＥＡと、撮像手段４０と、出力手段５０とで分散性判定システムＥＡ１を構成してもよいし、分散性判定装置ＥＡと通信可能な外部の入力装置からの操作信号を分散性判定装置ＥＡに入力してもよい。
The operation means 30 may be configured so as to be separable from the dispersibility determination device EA, or may be configured so as not to be separable from the dispersibility determination device EA.
The operation means 30 may or may not be included in the dispersibility determination device EA. If it is not included, the dispersibility determination system EA1 may be formed by the dispersibility determination device EA, the imaging means 40, and the output means 50, or an operation signal may be input to the dispersibility determination device EA from an external input device capable of communicating with the dispersibility determination device EA.
撮像手段４０は、分散性判定装置ＥＡに備わっていてもよいし、分散性判定システムＥＡ１に備わっていてもよいし、分散性判定装置ＥＡや分散性判定システムＥＡ１に備わっていなくてもよく、備わっていない場合、分散性判定装置ＥＡや分散性判定システムＥＡ１と通信可能な外部の撮像装置での撮像画像を画像取得手段２１で取得するようにしてもよい。 The imaging means 40 may be provided in the dispersibility determination device EA, or in the dispersibility determination system EA1, or may not be provided in the dispersibility determination device EA or the dispersibility determination system EA1. If it is not provided, the image acquisition means 21 may acquire an image captured by an external imaging device that can communicate with the dispersibility determination device EA or the dispersibility determination system EA1.
出力手段５０は、分散性判定装置ＥＡに備わっていてもよいし、分散性判定システムＥＡ１に備わっていてもよいし、分散性判定装置ＥＡや分散性判定システムＥＡ１に備わっていなくてもよく、備わっていない場合、分散性判定装置ＥＡや分散性判定システムＥＡ１と通信可能な外部の出力装置に判定結果を出力してもよい。 The output means 50 may be provided in the dispersibility determination device EA, may be provided in the dispersibility determination system EA1, or may not be provided in the dispersibility determination device EA or the dispersibility determination system EA1. If it is not provided, the determination result may be output to an external output device capable of communicating with the dispersibility determination device EA or the dispersibility determination system EA1.
学習済みモデルは、教師画像から特徴量を抽出しないで機械学習させたものであってもよく、例えば、教師画像からディープラーニングで学習させたものであってもよい。ディープラーニングで学習させる場合、特徴量抽出手段２２や特徴量抽出工程は、なくてもよい。
学習済みモデルの生成において、フローカーブ測定工程、フローカーブ判定工程および教師画像撮像工程の実施順序は特に限定されず、フローカーブ測定工程、教師画像撮像工程、フローカーブ判定工程の順に実施してもよいし、フローカーブ測定工程、フローカーブ判定工程、教師画像撮像工程の順に実施してもよいし、教師画像撮像工程、フローカーブ測定工程、フローカーブ判定工程の順に実施してもよい。
The trained model may be one that has been trained by machine learning without extracting features from a teacher image, or may be one that has been trained by deep learning from a teacher image. In the case of training by deep learning, the feature extraction means 22 and the feature extraction process may not be necessary.
In generating a trained model, the order in which the flow curve measurement process, flow curve determination process and teacher image capturing process are performed is not particularly limited, and they may be performed in the order of flow curve measurement process, teacher image capturing process and flow curve determination process, or they may be performed in the order of teacher image capturing process, flow curve measurement process and flow curve determination process.
粒子分散液は、分散媒が液体であれば、分散媒および分散質の物質、種別、組成等は、特に限定されることはない。例えば、分散媒は、有機溶剤、水、アルコール、樹脂溶液、油等であってもよい。また、分散質は、無機粒子、有機粒子、有機無機ハイブリッド粒子、マイクロカプセル、熱膨張粒子、または液滴等であってもよいし、粒子の表面が改質されたものであってもよいし、酸化物や水酸化物であってもよく、１種類の分散質が単独で用いられてもよいし、複数種類の分散質が併用されてもよい。 As long as the dispersion medium is a liquid, the substance, type, composition, etc. of the dispersion medium and dispersoid of the particle dispersion liquid are not particularly limited. For example, the dispersion medium may be an organic solvent, water, alcohol, a resin solution, oil, etc. The dispersoid may be inorganic particles, organic particles, organic-inorganic hybrid particles, microcapsules, thermally expandable particles, liquid droplets, etc., or may be particles with modified surfaces, oxides, or hydroxides. One type of dispersoid may be used alone, or multiple types of dispersoids may be used in combination.
本発明における手段および工程は、それら手段および工程について説明した動作、機能または工程を果たすことができる限りなんら限定されることはなく、まして、前記実施形態で示した単なる一実施形態の構成物や工程に全く限定されることはない。例えば、画像取得工程は、粒子分散液を撮像した撮像画像を取得する工程であればどのような工程でもよく、出願当初の技術常識に照らし合わせてその技術範囲内のものであればなんら限定されることはない（その他の手段および工程も同じ）。 The means and steps in the present invention are not limited in any way as long as they can perform the operations, functions or steps described for those means and steps, and are in no way limited to the components and steps of just one embodiment shown in the above embodiment. For example, the image acquisition step may be any step that acquires an image of a particle dispersion liquid, and is not limited in any way as long as it is within the technical scope in light of the common general technical knowledge at the time of filing (the same goes for the other means and steps).
ＥＡ…分散性判定装置
１０…記憶手段
２０…処理手段
２１…画像取得手段
２２…特徴量抽出手段
２３…分散性判定手段
３０…操作手段
４０…撮像手段
５０…出力手段
EA...
Claims (7)
前記粒子分散液を撮像した撮像画像を取得する画像取得工程と、
学習用の粒子分散液を撮像した教師画像と、前記学習用の粒子分散液について測定したフローカーブの形状特性から判定される前記分散性とを教師データとして機械学習させた学習済みモデルを用いて、前記撮像画像から前記分散性を判定する分散性判定工程とを実施することを特徴とする分散性判定方法。 A method for determining dispersibility of a particle dispersion, comprising the steps of:
an image acquisition step of acquiring an image of the particle dispersion;
a dispersibility determination process for determining the dispersibility from an image of a particle dispersion for training, using a trained model that has been machine-learned using a teacher image of the particle dispersion for training and the dispersibility determined from shape characteristics of a flow curve measured for the particle dispersion for training as training data.
前記撮像画像における粒子の数、最大粒子サイズ、および粒子の大きさの偏差を特徴量として抽出して前記分散性判定工程に引き渡す特徴量抽出工程を実施することを特徴とする請求項１に記載の分散性判定方法。 The trained model is machine-trained using the number of particles, the maximum particle size, and the deviation of particle size in the captured image as features,
The dispersibility determination method according to claim 1 , further comprising a feature extraction step of extracting the number of particles, the maximum particle size, and the deviation in particle size in the captured image as features and transferring them to the dispersibility determination step.
前記粒子分散液を撮像した撮像画像を取得する画像取得手段と、
学習用の粒子分散液を撮像した教師画像と、前記学習用の粒子分散液について測定したフローカーブの形状特性から判定される前記分散性とを教師データとして機械学習させた学習済みモデルを用いて、前記撮像画像から前記分散性を判定する分散性判定手段とを備えていることを特徴とする分散性判定装置。 A dispersibility determination device for determining the dispersibility of a particle dispersion liquid, comprising:
an image acquisition means for acquiring an image of the particle dispersion liquid;
a dispersibility determination means for determining the dispersibility from a captured image by using a trained model that has been machine-learned using a teacher image of a training particle dispersion and the dispersibility determined from shape characteristics of a flow curve measured for the training particle dispersion as training data.
学習用の粒子分散液のフローカーブを測定するフローカーブ測定工程と、
前記学習用の粒子分散液を撮像して教師画像とする教師画像撮像工程と、
前記フローカーブ測定工程で測定された前記フローカーブの形状特性から前記分散性を判定するフローカーブ判定工程と、
前記教師画像撮像工程で撮像された前記教師画像と、前記フローカーブ判定工程で判定された前記分散性とを教師データとして機械学習させ、前記学習済みモデルを生成するモデル生成工程とを実施することを特徴とする学習済みモデルの生成方法。 A method for generating a trained model for determining dispersibility of a particle dispersion, comprising:
a flow curve measuring step of measuring a flow curve of a particle dispersion for learning;
a teacher image capturing step of capturing an image of the particle dispersion liquid for learning to obtain a teacher image;
a flow curve determination step of determining the dispersibility from the shape characteristics of the flow curve measured in the flow curve measurement step;
A method for generating a trained model, comprising carrying out a model generation process in which the teacher image captured in the teacher image capturing process and the dispersibility determined in the flow curve determination process are machine-learned as teacher data to generate the trained model.
前記モデル生成工程では、前記特徴量抽出工程で抽出された粒子の数、最大粒子サイズ、および粒子の大きさの偏差を特徴量として機械学習させることを特徴とする請求項５に記載の学習済みモデルの生成方法。 carrying out a feature extraction step of extracting the number of particles, the maximum particle size, and the deviation of particle size in the teacher image as feature amounts;
The method for generating a trained model according to claim 5, characterized in that in the model generation step, the number of particles, the maximum particle size, and the deviation of particle sizes extracted in the feature extraction step are machine-learned as features.
The shape characteristic of the flow curve is a ratio of a viscosity when the shear rate in the flow curve is 0.1 s -1 to a viscosity when the shear rate is 10 s -1. The method for generating a trained model according to claim 5 or 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2021043856A JP7560393B2 (en) | 2021-03-17 | 2021-03-17 | Distributedness determination method, distributedness determination device, and trained model generation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2021043856A JP7560393B2 (en) | 2021-03-17 | 2021-03-17 | Distributedness determination method, distributedness determination device, and trained model generation method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| JP2022143379A JP2022143379A (en) | 2022-10-03 |
| JP7560393B2 true JP7560393B2 (en) | 2024-10-02 |
Family
ID=83454193
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP2021043856A Active JP7560393B2 (en) | 2021-03-17 | 2021-03-17 | Distributedness determination method, distributedness determination device, and trained model generation method |
Country Status (1)
| Country | Link |
|---|---|
| JP (1) | JP7560393B2 (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP7542290B1 (en) | 2024-03-26 | 2024-08-30 | 株式会社写真化学 | Mixing/defoaming processing data analysis device, processing condition determination device, and mixing/defoaming processing system |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2001313027A (en) | 2000-05-01 | 2001-11-09 | Ngk Insulators Ltd | Adjustment method of electrode slurry for lithium secondary battery |
| JP2007078590A (en) | 2005-09-15 | 2007-03-29 | Seishin Enterprise Co Ltd | Particle property analysis display device |
| JP2008292317A (en) | 2007-05-24 | 2008-12-04 | Sumitomo Chemical Co Ltd | Viscosity characteristic estimation method and viscosity characteristic estimation program |
| JP2011069754A (en) | 2009-09-28 | 2011-04-07 | Mitsubishi Rayon Co Ltd | Shearing viscosity estimation method and rheogram creation method |
| US20130084370A1 (en) | 2011-09-30 | 2013-04-04 | Krones Ag | Viscosity-controlled processing of liquid food |
| JP2017504007A (en) | 2013-12-09 | 2017-02-02 | テキサス テック ユニヴァーシティー システムＴｅｘａｓ Ｔｅｃｈ Ｕｎｉｖｅｒｓｉｔｙ Ｓｙｓｔｅｍ | Smartphone-based multiplexed viscometer for high-throughput analysis of fluids |
| WO2018104580A1 (en) | 2016-12-05 | 2018-06-14 | Valmet Automation Oy | Apparatus and method for measuring suspension and controlling process of suspension |
| WO2020175481A1 (en) | 2019-02-28 | 2020-09-03 | 地方独立行政法人神奈川県立産業技術総合研究所 | Fluid sample internal structure observation device and internal structure analysis system, fluid sample internal structure observation method and internal structure analysis method, and method for manufacturing ceramic |
| JP2020190508A (en) | 2019-05-23 | 2020-11-26 | 株式会社堀場製作所 | Particle size distribution measuring device, particle size distribution measurement method, and program for particle size distribution measuring device |
-
2021
- 2021-03-17 JP JP2021043856A patent/JP7560393B2/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2001313027A (en) | 2000-05-01 | 2001-11-09 | Ngk Insulators Ltd | Adjustment method of electrode slurry for lithium secondary battery |
| JP2007078590A (en) | 2005-09-15 | 2007-03-29 | Seishin Enterprise Co Ltd | Particle property analysis display device |
| JP2008292317A (en) | 2007-05-24 | 2008-12-04 | Sumitomo Chemical Co Ltd | Viscosity characteristic estimation method and viscosity characteristic estimation program |
| JP2011069754A (en) | 2009-09-28 | 2011-04-07 | Mitsubishi Rayon Co Ltd | Shearing viscosity estimation method and rheogram creation method |
| US20130084370A1 (en) | 2011-09-30 | 2013-04-04 | Krones Ag | Viscosity-controlled processing of liquid food |
| JP2017504007A (en) | 2013-12-09 | 2017-02-02 | テキサス テック ユニヴァーシティー システムＴｅｘａｓ Ｔｅｃｈ Ｕｎｉｖｅｒｓｉｔｙ Ｓｙｓｔｅｍ | Smartphone-based multiplexed viscometer for high-throughput analysis of fluids |
| WO2018104580A1 (en) | 2016-12-05 | 2018-06-14 | Valmet Automation Oy | Apparatus and method for measuring suspension and controlling process of suspension |
| WO2020175481A1 (en) | 2019-02-28 | 2020-09-03 | 地方独立行政法人神奈川県立産業技術総合研究所 | Fluid sample internal structure observation device and internal structure analysis system, fluid sample internal structure observation method and internal structure analysis method, and method for manufacturing ceramic |
| JP2020190508A (en) | 2019-05-23 | 2020-11-26 | 株式会社堀場製作所 | Particle size distribution measuring device, particle size distribution measurement method, and program for particle size distribution measuring device |
Also Published As
| Publication number | Publication date |
|---|---|
| JP2022143379A (en) | 2022-10-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Bourcier et al. | Influence of particle size and shape properties on cake resistance and compressibility during pressure filtration | |
| Maaß et al. | Automated drop detection using image analysis for online particle size monitoring in multiphase systems | |
| Dinkgreve et al. | Carbopol: From a simple to a thixotropic yield stress fluid | |
| Kacker et al. | Application of inline imaging for monitoring crystallization process in a continuous oscillatory baffled crystallizer | |
| Chan et al. | Particle sizing of nanoparticle adjuvant formulations by dynamic light scattering (DLS) and nanoparticle tracking analysis (NTA) | |
| Abidin et al. | Review on measurement techniques for drop size distribution in a stirred vessel | |
| Panckow et al. | Photo-optical in-situ measurement of drop size distributions: Applications in research and industry | |
| US11080512B2 (en) | Information processing device, information processing method, measurement system and non-transitory storage medium | |
| JP7418639B2 (en) | Particle analysis data generation method, particle analysis data generation program, and particle analysis data generation device | |
| Muhaimin et al. | Real-time particle size analysis using focused beam reflectance measurement as a process analytical technology tool for continuous microencapsulation process | |
| Mickler et al. | Drop swarm analysis in dispersions with incident-light and transmitted-light illumination | |
| WO2021102844A1 (en) | Method, device and system for processing image | |
| JP7560393B2 (en) | Distributedness determination method, distributedness determination device, and trained model generation method | |
| CN114460005A (en) | Dynamic microscopic imaging red blood cell multi-angle morphology measuring method and system | |
| Afik et al. | On the role of initial velocities in pair dispersion in a microfluidic chaotic flow | |
| CN111912745A (en) | A method for measuring liquid viscosity by drop experiment | |
| Burke et al. | Two deep learning methods in comparison to characterize droplet sizes in emulsification flow processes | |
| Muhaimin et al. | Polymer type effect on PLGA-based microparticles preparation by solvent evaporation method with single emulsion system using focussed beam reflectance measurement | |
| Lecanu et al. | Predicting the gravity-driven flow of power law fluids in a syringe: a rheological map for the IDDSI classification | |
| Caserta et al. | Evolution of drop size distribution of polymer blends under shear flow by optical sectioning | |
| Mozafari et al. | An automated platform for measuring infant formula powder rehydration quality using a collaborative robot integrated with computer vision | |
| Mehri et al. | Controlled microfluidic environment for dynamic investigation of red blood cell aggregation | |
| Srivastava et al. | Bubble characterization in a continuous casting mold: comparison and identification of image processing techniques | |
| Bran et al. | Fluid rheology prediction using interface detection and machine learning regression | |
| Makni et al. | A new method of mixing quality assessment for friction material constituents toward better mechanical properties |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| A621 | Written request for application examination |
Free format text: JAPANESE INTERMEDIATE CODE: A621 Effective date: 20240112 |
|
| A977 | Report on retrieval |
Free format text: JAPANESE INTERMEDIATE CODE: A971007 Effective date: 20240826 |
|
| TRDD | Decision of grant or rejection written | ||
| A01 | Written decision to grant a patent or to grant a registration (utility model) |
Free format text: JAPANESE INTERMEDIATE CODE: A01 Effective date: 20240910 |
|
| A61 | First payment of annual fees (during grant procedure) |
Free format text: JAPANESE INTERMEDIATE CODE: A61 Effective date: 20240919 |
|
| R150 | Certificate of patent or registration of utility model |
Ref document number: 7560393 Country of ref document: JP Free format text: JAPANESE INTERMEDIATE CODE: R150 |