AR tooltip tracking raw data and processing codes

Published: 13 July 2026| Version 2 | DOI: 10.17632/jxww2xj2dr.2
Contributor:
Federico Salerno

Description

DATASET OVERVIEW This dataset contains the raw data collected from tooltip tracking using both Microsoft HoloLens 2 and a stereo-based system that employs two RealSense SR305 RGB cameras. It also includes ground truth reference points and a Python script designed for processing the raw data. FOLDER STRUCTURE The data input folder is organized as follows: input_points/ │ ├── qualified_points.txt │ └── Measured/ ├── HoloLens_participant1/ │ ├── measurement_1.txt │ ├── measurement_2.txt │ └── ... (other measured files) └── Realsense_operatore3/ ├── measurement_1.txt ├── measurement_2.txt └── ... (other measured files) measurement_artifact_STL/ ├── Artifact.STL ├── ArUco_adapter.STL └── ArUco_adapter_female.STL CONTENTS qualified_points.txt: This file contains the ground truth reference points, which serve as the benchmark for aligning and comparing the measured point clouds. Each line in the file represents a 3D point in space, formatted as X, Y, and Z coordinates. Measured/: This directory houses the measured 3D point clouds, which are categorized based on the device or participant involved in the data collection process. The subdirectories within this folder correspond to specific sessions or devices e.g.: HoloLens_participant1/: Contains data collected using the Microsoft HoloLens 2 from participant 1. RealSense_partecipant3/: Contains data collected using the stereo-based system with two RealSense SR305 cameras from participant 3. measurement_artifact_STL/: This folder contains the 3D models (STL format) of the measurement artifact, so that it can be 3D printed and reproduced. Artifact.STL is the main measurement artifact body; ArUco_adapter.STL; ArUco_adapter_female.STL are the male and female parts of the adapter used to mount the ArUco marker on the artifact. PYTHON SCRIPT local_coordinate_differences.py: This script first creates a local reference coordinate system based on three points. It then transforms the measured 3D points relative to this local reference system. Finally, it calculates the differences between the transformed points and the ground truth reference points from the qualified points set.

Files

Institutions

Categories

Computer Vision, Image Processing, Augmented Reality

Licence