JP2020008916A5 - - Google Patents
Download PDFInfo
- Publication number
- JP2020008916A5 JP2020008916A5 JP2018126596A JP2018126596A JP2020008916A5 JP 2020008916 A5 JP2020008916 A5 JP 2020008916A5 JP 2018126596 A JP2018126596 A JP 2018126596A JP 2018126596 A JP2018126596 A JP 2018126596A JP 2020008916 A5 JP2020008916 A5 JP 2020008916A5
- Authority
- JP
- Japan
- Prior art keywords
- measurement data
- abundance
- usefulness
- rotation
- feature amount
- 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.)
- Granted
Links
- 238000005259 measurement Methods 0.000 claims description 27
- 238000001514 detection method Methods 0.000 claims description 24
- 230000001419 dependent Effects 0.000 claims 9
- 239000000284 extract Substances 0.000 claims 3
- 238000000605 extraction Methods 0.000 claims 1
- 238000000034 method Methods 0.000 claims 1
Description
（４）本発明に係る学習装置は、上記（１）に記載の物体検出装置に関する学習に用いる学習装置であって、学習用の前記計測データ、及び当該学習用の前記計測データに関する前記存在度の教師データを取得する手段と、前記学習用の前記計測データを前記物体検出装置に入力する入力手段と、前記学習用の前記計測データに対して前記存在度算出手段が算出した前記存在度と、前記教師データとを比較する比較手段と、前記比較の結果に基づいて、前記有用度比率決定手段にて前記有用度比率を定める算出器のパラメータを変更するパラメータ変更手段と、を備える。 (4) the learning apparatus according to the present invention, there is provided a learning apparatus used for learning about the object detection apparatus according to (1), the measurement data, and the abundance relating to the measurement data for the learning of the learning means for acquiring training data of an input means for inputting the measurement data for the learning to the object detecting device, and the abundance of the existing degree calculating unit is calculated for the measurement data for the learning A comparison means for comparing the teacher data and a parameter changing means for changing the parameters of the calculator for determining the usefulness ratio by the usefulness ratio determining means based on the result of the comparison.
Claims (8)
前記計測データに基づいて、当該計測データから抽出される、前記物体の回転角度に対する依存性が高い回転依存量と前記依存性が低い回転不変量との有用度の比率を定める有用度比率決定手段と、
前記計測データから、前記有用度比率決定手段が定めた有用度比率に従って前記回転依存量及び前記回転不変量の少なくとも一方を含んだ検出用特徴量を抽出する特徴量抽出手段と、
前記検出用特徴量を入力されて前記計測データに前記物体が現れている度合いを表す存在度を算出する存在度算出手段と、
前記存在度に基づいて前記物体を検出する物体検出手段と、
を備えたことを特徴とする物体検出装置。 An object detection device that detects a predetermined object from the measurement data to be processed.
Usefulness ratio determining means for determining the ratio of usefulness between a rotation-dependent amount having a high dependence on the rotation angle of the object and a rotation invariant having a low dependence, which is extracted from the measurement data based on the measurement data. When,
A feature amount extracting means for extracting a detection feature amount including at least one of the rotation-dependent amount and the rotation invariant amount from the measurement data according to the usefulness ratio determined by the usefulness ratio determining means.
An abundance calculation means for inputting the detection feature amount and calculating the abundance representing the degree of appearance of the object in the measurement data.
An object detection means that detects the object based on the abundance, and
Object detecting apparatus comprising the.
前記特徴量抽出手段は、前記複数の局所領域ごとに前記検出用特徴量を抽出し、
前記存在度算出手段は、前記複数の局所領域それぞれについて、前記検出用特徴量から前記物体の部位について部位存在度を算出し、当該部位存在度を統合して前記存在度を求めること、
を特徴とする請求項１から請求項３のいずれか１つに記載の物体検出装置。 The usefulness ratio determining means determines the usefulness ratio for each of a plurality of local regions in the space from which the measurement data has been acquired.
The feature amount extracting means extracts the detection feature amount for each of the plurality of local regions, and then extracts the feature amount for detection.
The abundance calculation means calculates the abundance of a part of the object from the feature amount for detection for each of the plurality of local regions, and integrates the abundance of the part to obtain the abundance.
The object detection device according to any one of claims 1 to 3.
前記計測データに基づいて、当該計測データから抽出される、前記物体の回転角度に対する依存性が高い回転依存量と前記依存性が低い回転不変量との有用度の比率を定める有用度比率決定手段、
前記計測データから、前記有用度比率決定手段が定めた有用度比率に従って前記回転依存量及び前記回転不変量の少なくとも一方を含んだ検出用特徴量を抽出する特徴量抽出手段、
前記検出用特徴量を入力されて前記計測データに前記物体が現れている度合いを表す存在度を算出する存在度算出手段、及び、
前記存在度に基づいて前記物体を検出する物体検出手段、
として機能させることを特徴とする物体検出プログラム。 A program for causing a computer to perform a process of detecting a predetermined object from the measurement data to be processed.
Usefulness ratio determining means for determining the ratio of usefulness between a rotation-dependent amount having a high dependence on the rotation angle of the object and a rotation invariant having a low dependence, which is extracted from the measurement data based on the measurement data. ,
A feature amount extracting means for extracting a detection feature amount including at least one of the rotation-dependent amount and the rotation invariant amount from the measurement data according to the usefulness ratio determined by the usefulness ratio determining means.
An abundance calculation means for inputting the detection feature amount and calculating the abundance indicating the degree to which the object appears in the measurement data, and
An object detection means that detects the object based on the abundance,
An object detection program characterized by functioning as.
前記計測データに基づいて、当該計測データから抽出される、前記物体の回転角度に対する依存性が高い回転依存量と前記依存性が低い回転不変量との有用度の比率を定める有用度比率決定ステップと、
前記計測データから、前記有用度比率決定ステップにて定めた有用度比率に従って前記回転依存量及び前記回転不変量の少なくとも一方を含んだ検出用特徴量を抽出する特徴量抽出ステップと、
前記検出用特徴量を入力されて前記計測データに前記物体が現れている度合いを表す存在度を算出する存在度算出ステップと、
前記存在度に基づいて前記物体を検出する物体検出ステップと、
を備えたことを特徴とする物体検出方法。 An object detection method that detects a predetermined object from the measurement data to be processed.
Usefulness ratio determination step for determining the ratio of usefulness between a rotation-dependent amount having a high dependence on the rotation angle of the object and a rotation invariant having a low dependence extracted from the measurement data based on the measurement data. When,
A feature amount extraction step for extracting a detection feature amount including at least one of the rotation-dependent amount and the rotation invariant amount from the measurement data according to the usefulness ratio determined in the usefulness ratio determination step.
The abundance calculation step of inputting the detection feature amount and calculating the abundance representing the degree of appearance of the object in the measurement data, and the abundance calculation step.
An object detection step that detects the object based on the abundance,
Object detecting method characterized by comprising a.
学習用の前記計測データ、及び当該学習用の前記計測データに関する前記存在度の教師データを取得する手段と、
前記学習用の前記計測データを前記物体検出装置に入力する入力手段と、
前記学習用の前記計測データに対して前記存在度算出手段が算出した前記存在度と、前記教師データとを比較する比較手段と、
前記比較の結果に基づいて、前記有用度比率決定手段にて前記有用度比率を定める算出器のパラメータを変更するパラメータ変更手段と、
を備えることを特徴とする学習装置。 A learning device used for learning about the object detection device according to any one of claims 1 to 5.
The measurement data for learning, and means for acquiring training data of the abundance for said measurement data for the learning,
An input means for inputting the measurement data for learning into the object detection device, and
It said abundance of the existing degree calculating unit is calculated for the measurement data for the learning, and comparing means for comparing the teacher data,
Based on the result of the comparison, the parameter changing means for changing the parameter of the calculator that determines the usefulness ratio by the usefulness ratio determining means, and the parameter changing means.
A learning device characterized by being provided with.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2018126596A JP7128578B2 (en) | 2018-07-03 | 2018-07-03 | Object detection device, object detection program, object detection method, and learning device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2018126596A JP7128578B2 (en) | 2018-07-03 | 2018-07-03 | Object detection device, object detection program, object detection method, and learning device |
Publications (3)
| Publication Number | Publication Date |
|---|---|
| JP2020008916A JP2020008916A (en) | 2020-01-16 |
| JP2020008916A5 true JP2020008916A5 (en) | 2021-07-26 |
| JP7128578B2 JP7128578B2 (en) | 2022-08-31 |
Family
ID=69151767
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP2018126596A Active JP7128578B2 (en) | 2018-07-03 | 2018-07-03 | Object detection device, object detection program, object detection method, and learning device |
Country Status (1)
| Country | Link |
|---|---|
| JP (1) | JP7128578B2 (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022180870A1 (en) * | 2021-02-26 | 2022-09-01 | 日本電気株式会社 | Learning device, learning method, and recording medium |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2012133445A (en) | 2010-12-20 | 2012-07-12 | Panasonic Corp | Target detection apparatus and target detection method |
-
2018
- 2018-07-03 JP JP2018126596A patent/JP7128578B2/en active Active
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP6610678B2 (en) | Peak detection method and data processing apparatus | |
| US20200150159A1 (en) | Anomaly detection device, anomaly detection method, and storage medium | |
| WO2016155241A1 (en) | Method, system and computer device for capacity prediction based on kalman filter | |
| CN113392742A (en) | Abnormal action determination method and device, electronic equipment and storage medium | |
| CN109901996B (en) | Auxiliary test method and device, electronic equipment and readable storage medium | |
| JP2010134649A5 (en) | ||
| CN104376319B (en) | A kind of method based on anisotropic Gaussian core extraction closed edge image outline | |
| JP2009288027A5 (en) | ||
| JP2019057173A5 (en) | ||
| KR20170053692A (en) | Apparatus and method for ensembles of kernel regression models | |
| CN111207926A (en) | Fault diagnosis method based on rolling bearing, electronic device and storage medium | |
| JP2019046094A5 (en) | ||
| JP2015032001A5 (en) | ||
| WO2015152181A1 (en) | Measurement value analysis device and measurement value analysis method | |
| CN109582988B (en) | Method for determining position of strain gauge for monitoring vibration stress of blade of aircraft engine | |
| JP2018036528A5 (en) | ||
| JP2020008916A5 (en) | ||
| JP2011243007A5 (en) | ||
| CN107561106B (en) | Method and device for measuring characterization parameters of streak-shaped morphology | |
| JP2015085437A (en) | Determination device, determination method, and determination program | |
| WO2021130978A1 (en) | Motion analysis system and motion analysis program | |
| JPWO2021144938A5 (en) | ||
| JP2022013433A5 (en) | ||
| CN113239075A (en) | Construction data self-checking method and system | |
| CN109993022B (en) | Height detection method and method for establishing height detection equation |