measurementのメーカーや取扱い企業、製品情報、参考価格、ランキングをまとめています。
イプロスは、 製造業 BtoB における情報を集めた国内最大級の技術データベースサイトです。

measurement×HALデザイン研究所 - メーカー・企業と製品の一覧

measurementの製品一覧

1~2 件を表示 / 全 2 件

表示件数

Motion measurement using Azure Kinect

We conducted walking measurements using two Azure Kinect devices. Please make use of the insights gained from this process.

【Motion Measurement with Azure Kinect】 Our company has successfully utilized the Azure Kinect, the successor to Kinect 2, with our technology, achieving the potential to use it as a markerless 3D motion measurement system. We have already published an overview of the insights gained during this process, so please refer to our catalog. This time, we conducted actual walking measurements. Below, we outline the issues encountered during the experiments and the countermeasures taken, as well as examples of walking measurement results, which are introduced in the catalog. If you are interested, please contact us. 【Issues Encountered】 Errors may occur during the startup process when powering on the Azure Kinect, depending on the conditions of the capture range, and it may become impossible to recover from these errors. This happens when there are objects within the field of view that the Azure Kinect cannot measure. Specifically, this includes objects that are either too close or too far away. Errors also occur when the sky is within the field of view. 【Countermeasures】 It is necessary to follow quite complex procedures to recover from this issue. If you are having trouble with this matter, please contact us.

  • 3D measuring device

ブックマークに追加いたしました

ブックマーク一覧

ブックマークを削除いたしました

ブックマーク一覧

これ以上ブックマークできません

会員登録すると、ブックマークできる件数が増えて、ラベルをつけて整理することもできます

無料会員登録

Markerless 3D measurement of human joint points using OpenPose.

We will introduce a method to obtain three-dimensional data using two-dimensional joint point data measured by OpenPose, without markers.

**What is OpenPose?** OpenPose is a method developed by Zhe Cao and others at Carnegie Mellon University (CMU) that visualizes human poses using deep learning. Specifically, it can detect human joint points simply by inputting a still image. In other words, displaying the positions of joints as points makes it easier to understand angles at the knees or the degree of elbow bending. However, it is important to note that these are purely 2D data and do not take depth into account. **Expansion from 2D to 3D** By utilizing our proprietary calibration technology, we can obtain markerless 3D surface point cloud data of a person. We can then apply the 2D joint point coordinates obtained from OpenPose to this data to derive 3D joint point coordinates. **Number of Cameras and Measurement Accuracy** - With 3 cameras: We can measure the full-body surface point cloud data, allowing for accurate joint point coordinates. - With 2 cameras: We cannot measure the full-body surface point cloud data. - With 1 camera: Please refer to the attached materials for more details.

  • 3D measuring device

ブックマークに追加いたしました

ブックマーク一覧

ブックマークを削除いたしました

ブックマーク一覧

これ以上ブックマークできません

会員登録すると、ブックマークできる件数が増えて、ラベルをつけて整理することもできます

無料会員登録