1. Recent developments on micro-triangulation
for fiducial points and wires
Vasileios Vlachakis

2. To link the fiducials with the wire, in geometric sense.
@ François Nicolas Morel, CERN
To link the fiducials with the wire, in geometric sense.
Mechanical axis → fiducials
Magnetic axis stretched
Electric axis wire
To measure the fiducials and the wire using theodolites.
Observations → angles
(horizontal, vertical)
Leica TDA5005 theodolite.
QDaedalus measuring system.
Computer vision techniques.
Least-squares analysis.

3. Copper-Beryllium (CuBe) wire: Ø 100 μm.
@ François Nicolas Morel, CERN
Leica TDA5005:
Robotic theodolite.
High accuracy:
0.5 arcsec μm m
Copper-Beryllium (CuBe) wire:
Ø 100 μm.
Ceramic spheres:
0.5 inch
Grade 40 (sphericity 1 μm)

4. Main features  Advantages: Precise measurements – few micrometers
Contactless measurements – ideal for stretched wire
Automatic measurements – no need of observer
Remote-controlled measurements – 7 m meters in cable, ∞ on-line
Fast measurements – a series in <10 min (depending on the setup, conditions, etc.)
Portable – few boxes…
Main features
♡ For fiducialization:
Introduces the direct stretched wire observation (can lead to various applications related to alignment techniques).
It is expected to have lower precision than a CMM.
It is a low-cost, portable solution based on standard surveying-geodetic techniques.
 Disadvantage:
The scale should be introduced using coordinates or distances as geodetic network constraints.

5. QDaedalus Developed at ETH Zurich Consists of hardware and software
Based on
industrial robotic theodolites
Reversible replacement
eye-piece ↔ CCD camera
Uses Optical Target Recognition
QDaedalus
Overview
The QDaedalus is going to be used in this study. The system in developed at ETH Zurich.
QDaedalus is based on industrial robotic theodolites equipped with external CCD camera in a reversible way.
Using Optical Target Recognition algorithms (explained later), the system can extract and measure various kinds od targets.
B. Bürki et al., “DAEDALUS: A versatile usable digital clip-on measuring system for Total Stations”, in IPIN2010, Zurich, Switzerland, 2010.

6. OTR Algorithms 1 2 QDaedalus 3 4 5 6 1. Centre of mass
2. Template least-squares matching 1 2
3. Ellipse matching
QDaedalus
Optical Target Recognition
4. Circle matching
5. Line matching
6. Multi-Ellipse matching 3 4 5 6
The optical target recognition algorithms.
Centre of mass: used mainly for astro-geodetic applications
Template least squares matching: used only in the calibration process
Ellipse matching: used on circular targets, e.g. photogrammetric targets
Circle matching: used on spherical targets
Line matching: used on the stretched wire (developed in this study)
Multi-Ellipse matching: used on existing circular targets during dam monitoring
S. Guillaume et al., “QDaedalus: Augmentation of Total Stations by CCD Sensor for Automated Contactless High-Precision Metrology”, in FIG Working Week 2012, Rome, Italy, 2012.

7. Aim: To measure the targets (fiducials) and
the stretched wire in a common coordinate system.
To validate the measuring system.
To develop and validate a wire detection and measurement algorithm.
To develop an integrated geodetic network including spherical targets, stretched wire reconstruction, theodolite systematic errors.
To simulate measurement configurations for precision and efficiency optimization.
Tasks
This project aims to develop:
Accurate and robust algorithm in C++ for the detection and measurement of the stretched wire
Least-squares analysis algorithm that includes the systematic errors of the theodolite in the mathematical model
Integrated geodetic network with both the spherical targets (fiducials) and the stretched wire reconstruction (straight line or catenary shape)
Simulation of the measurement configuration for the optimisation of the measurement

8. Factors that may affect the measurement
Thermal effects
Camera, theodolite sensors, tripods
System validation

9. Measurement Stability

10. Measurement Stability
10 min
60 min
30 min
230 min
Measurement Stability
50 min
100 min
50 min
50 min
100 min

11. Factors that may affect the measurement
Thermal effects
Camera, theodolite sensors, tripods
Camera calibration
Distance: theodolite ↔ calibration target
Distance: calibration target ↔ target
Type of calibration target
System validation

12. Calibration

13. Factors that may affect the measurement
Thermal effects
Camera, theodolite sensors, tripods
Camera calibration
Distance: theodolite ↔ calibration target
Distance: calibration target ↔ target
Type of calibration target
Focus on the target
Algorithm (circle, line)
Range of focus VS distance
System validation

14. Focus

15. Factors that may affect the measurement
Thermal effects
Camera, theodolite sensors, tripods
Camera calibration
Distance: theodolite ↔ calibration target
Distance: calibration target ↔ target
Type of calibration target
Focus on the target
Algorithm (circle, line)
Range of focus VS distance
Light conditions
Ambient, external, internal
External light effect VS distance
System validation

16. Wire detection and measurement algorithm
User defined parameters
Minimum edge points (for each edge): increases the robustness.
Width of Region Of Interest (ROI) window: increases precision avoiding the out-of-focus area.
Height of Region Of Interest (ROI) window: helps when the wire is depicted inclined.
Maximum residual: increases precision discarding outlier pixels w.r.t. the fitting line.
Canny threshold: increases the robustness in case of noisy images.

17. Wire detection and measurement algorithm

18. Angle observations’
fan

19. PACMAN bench configuration
theodolite stations
ceramic spheres
stretched wire
PACMAN bench configuration

20. PACMAN bench configuration
Observations to the targets
PACMAN bench configuration

21. PACMAN bench configuration
Observations to the wire
PACMAN bench configuration

22. PACMAN bench configuration

23. Least-squares analysis software
Input:
horizontal and vertical angles to targets and wires (with uncertainties).
Approximate coordinates of the network (with uncertainties).
datum constraints.
user’s parameters.
Output:
stations: coordinates, orientations, and 3 systematic errors per instrument.
targets: coordinates.
wires: position, orientation and coordinates of the observed points.
uncertainties.
Least-squares analysis software

24. Tasks:
Validation of the measuring system.
Development and validation of a wire detection and measurement algorithm.
Development of an integrated geodetic network.
Simulation of the measurement configurations.
Conclusion and outlook
Thank you!