Z+F Imager 5016 TLS Distance Uncertainty Benchmark Dataset

## Overview
This dataset provides high-precision TLS measurements from the Z+F Imager 5016 laser scanner, referenced against laser trackers (LT). It enables the analysis and modeling of TLS distance uncertainties, including both systematic deviations and precision. The dataset is intended for research on TLS calibration, uncertainty estimation, and machine learning-based modeling (see Fig. 1 for an overview of the experimental setup and features).

- **Scanner type:** Z+F Imager 5016 (two units with different *manufacturer calibrations*: TLS1 (Data set A & B) – October 2021, TLS2 (Data set C) – February 2023; cf. Fig. 1)
- **Reference systems:** Leica ATS600, Leica ATR960 (cf. Fig. 1)
- **Environment:** Controlled indoor laboratory (~20°C)
- **Total points:** ~21 million TLS points across all data sets
- **Data organization:** Separate files per panel; five columns per file (Intensity, Distance, Incidence angle, distance deviation, Combination ID)

## Data Sets Concept
- **Data set A (TLS1, Primary data set):** Contains TLS measurements from one scanner with high-precision reference. Provides a comprehensive basis for model training. *Acquired 13–16 June 2025.*
- **Data set B (TLS1, Repeat measurements):** Acquired with the same scanner at a different time, using a high-precision reference. Enables evaluation of temporal stability and repeatability of measurements. *Acquired 06–12 May 2025.*
- **Data set C (TLS2, Cross-device):** Acquired with a second scanner under similar conditions and reference as Data set A. Allows assessment of device-specific variations and generalization of models. *Acquired 13–16 June 2025.*

## Measurement Concept
- Measurements were performed using flat Alucore panels with varying surface reflectivities (0–86% white surface ratio; cf. Fig. 1).
- Panels were scanned at multiple distances and orientations (incidence angles).
- High-precision reference measurements were acquired with the LTs.
- Each TLS scan is transformed into the LT coordinate system for direct comparison.
- Backward modeling between LT and TLS measurements provides the TLS distance deviations.
- The dataset enables analysis of TLS measurement uncertainties across distance, incidence angle, and surface reflectivity.

## Data Structure
- Each dataset is organized by panel.
- **Columns (per row):**
1. Intensity
2. Distance
3. Incidence angle
4. Distance deviations (relative to reference)
5. Combination ID (links distance, incidence angle, and panel configuration)

- Files are named systematically to indicate the panel and data set.
- This structure allows flexible use for ML training, temporal analysis, and cross-device evaluation.

**Figure 1:** Experimental setup showing the TLS and reference sensors (top left), Alucore panels with varying reflectivities (top right), and the definition of the features used in the dataset (bottom).