Basics, principles and application of 3D imaging systems with conventional and high-resolution cameras

1
Basics, principles and application of 3D imaging
systems with conventional and high-resolution
cameras
Andreas Gaich 3G Software Measurement GmbH
Markus Pötsch Wulf Schubert Institute for Rock
Mechanics and Tunnelling Graz University of
Technology
Prepared forWorkshop on the Use of Modern
Technologies for Rock Face CharacterizationGolden
Rocks Conference 2006, Golden, CO
2
MOTIVATION
3
CURRENT PRACTICE
Evident subjectivity Access needed for
measurement Possibly working in hazardous
areas Incomplete, incorrect data Pressure of time
4
CURRENT PRACTICE
Access needed for measurement Incomplete
data Possibly working in hazardous areas Overview
problem
5
WHAT WAS THE VISION.

• Data acquisition in engineering geology
• requires manual access
• requires time
• is reproducible?
• WANTED
• Economic system for extensive measurements and
  thorough assessments of geometric rock mass
  properties

? Solution 3D images
6
3D IMAGES
What is a 3D image?
90 Megapixel 1.6 million measuring points
7
3D IMAGES
3D images for Geometrical, Geological und
Geotechnical assessments
Part of a marble quarry 10 benches 150 m
height 90 Megapixel 1.6 million measuring points
8
BASICS AND PRINCIPLES
9
STEREOSCOPIC GEOMETRY
10
REQUIREMENTS

• Two images showing the same object from different
  angles
• 2D Image coordinates
• Corresponding points in the two images
• Orientation between the projection planes of the
  cameras
• Internal projective mechanism of the cameras
• Calibration
• Scale information of the stereoscopic geometry

11
APPROACH

• DIGITAL PHOTOGRAPHY
• Pixel coordinates ? 2D Image coordinates
• Automatic image matching provides corresponding
  points
• COMPUTER VISION
• Relative image orientation
• 3D point reconstruction
• Camera calibration

12
GENERATION OF A GENERIC 3D IMAGE
13
MATCHING

• REQUIREMENTS ON MATCHING
• Detailed geometry requires a high amount of
  corresponding points
• This can only be done by an automatic algorithm
• Robust algorithm to prevent from mismatches

14
MATCHING POINT DENSITY
37 cm/point 1000
16 cm/point 5000
2.5 cm/point 200000
Density No. of points
Height 10m
15
3D RECONSTRUCTION
3D image reconstruction from corresponding points
and determined relative orientation
Stereoscopic image pair
16
RESULTS
3D points
Complete, dense acquisition of object surface
17
FROM A GENERIC TO A METRIC 3D IMAGE
18
METRIC 3D IMAGE

• Issues for accurate measurements from a 3D image
• Calibration of the imaging system
• Individual calibration
• Pre-calibrated system
• Scaling or referencing the 3D image
• A scaled 3D image allows measurements in local
  coordinates
• A referenced allows measurements in a superior
  coordinate system

19
CALIBRATION

• Calibration means to determine the internal
  orientation of an imaging system described by
  several parameters, such as focal length, image
  centre, pixel size, and lens distortion.
• In other words how 3D world is projected onto
  the 3D image
• Calibration is usually performed by comparing the
  geometry of a known object with its image.
• Computer vision allows the application of 2D
  calibration objects instead of 3D objects.

20
CALIBRATION

• Calibration pattern

21
CALIBRATION

• Distortions of the imaging system

22
SCALING AND REFERENCING

• In order to obtain a metric 3D image
• objects with known size and/or orientation
• reference points with known co-ordinates are used
  

23
REFERENCING

• Range pole Reference point

24
3Gs3D IMAGING SYSTEMS
25
3D IMAGING SYSTEMS
PanoramaScanner Rotating high-resolution
line-scan camera
Calibrated SLR camera Conventional off-the-shelf
camera
26
3D IMAGING SYSTEMS
Components Zoom lens 20 300 mm Reference
targets Calibration target Control software 100
Megapixel 48 Bit Color depth Individual
calibration Number of 3D points typ. 700,000
1,600,000
Components Zoom lens 10 70 mm (typ.) Range
pole Calibration target Control software 6
Megapixel 24 Bit Color depth Pre-calibrated
system Number of 3D points typ. 100,000
200,000
27
3D IMAGING SYSTEMS
Horizontal field of view 360 Vertical field of
view according to lens Referencing using
surveyed reference targets Accuracy 2-3 cm
absolute error at 100-1000m distance
Horizontal and vertical field of view are coupled
and depend on the lens Joining of parts
possible Referencing using range pole in local
coordinates or surveyed reference targets
Accuracy depending on object size and
application typ. cm range
28
3D IMAGING SYSTEMS
Typical fields of application Large rock walls (lt
300 m) Mine pits Highly detailed images Large
distance applications (lt 1500 m)
Typical fields of application Bench
faces Outcrops (lt 30 m height) Tunnel faces
29
SITE PROCEDURE
30
SITE PROCEDURE

• 1. Setting up the range pole

31
SITE PROCEDURE

• 2. Taking left image (freely)

32
SITE PROCEDURE

• 3. Taking right image (freely)

33
SITE PROCEDURE

• 4. Generation of the 3D image on the PC

34
RESULT
Results 3D image Topography image information
35
SCALING

• Scaling

36

REFERENCING

• Algorithms
• Bundle Adjustment
• Direct Linear Transformation
• Levenberg-Marquardt

37
RESULTS
38
RESULTS

• 3D rock mass parameters

• GEOTECHNICAL PARAMETERS
• Positions (X,Y,Z) m
• Distances, lengths m
• Areas m2
• Orientations from faces /
• Orientations from traces /
• Bridges
• Free grouping to sets
• Spacing, Stereograms
• Export directly into CAD, MS Excel

39
RESULTS

• Geotechnical evaluations

Integrated determination of rock mass parameters
Spacing, Frequency,Fisher constant Hemispherical
plots
40
GEOMETRY OF THE BENCH FACE
?Rock wall height and width ?Overall
inclination ?Dense measurements over the whole
area ?Profiles
41
RESULTS

• Merging of single 3D images

42
TUNNELLING

• Reproducible and objective documentation
• More time for tunnel face assessment
• Excavation optimisation

43
APPLICATIONROCK SLOPES
44
ROCK SLOPE
45
ROCK SLOPE

• Assessment of joint sets
• Nor access problem

46
SPACING
47
STEREONET
48
ROCK SLOPE
49
APPLICATIONROCK SLOPES II
50
ROCK SLOPES

• Kinematic analysis of a rock block
• Data acquisition using JointMetriX3D
• Distance 700 m
• Imaging area 120 m x 420 m

51
ROCK SLOPES
Stability and Safety
52
ROCK SLOPES
Stability and Safety
53
ROCK SLOPES
Stability and Safety
54
ROCK SLOPES
Stability and Safety
Contact-free measurements
55
ROCK SLOPES
Stability and Safety
Contact-free measurements
56
ROCK SLOPES
Stability and Safety
Measurement of - lengths, distances -
orientations from traces
346.3 / 67.4 18,33 m
Contact-free measurements
57
ROCK SLOPES
Stability and Safety
346.3 / 67.4 18,33 m
Contact-free measurements
58
ROCK SLOPES
Stability and Safety
Measurement of - areas - orientations from
surfaces
346.3 / 67.4 18,33 m
232.6 / 66.4 71,04 m2

• Volume of removable block 303 m3

Contact-free measurements
59
ROCK SLOPES
Stability and Safety
346.3 / 67.4 18,33 m
232.6 / 66.4 71,04 m2

• Volume of removable block 303 m3

Contact-free measurements
60
ROCK SLOPES
Stability and Safety
346.3 / 67.4 18,33 m
232.6 / 66.4 71,04 m2

• Volume of removable block 303 m3

Contact-free measurements
61
ROCK SLOPES
FAILURE MODES
KINEMATICAL MOVABILITY Translation
Removability Rotation Rotatability
Falling/Lifting Single face sliding Double face
sliding
Corner rotation Edge rotation
62
APPLICATIONTUNNELLING
63
TUNNELLING

• Tunnelling

Data acquisition on site in a minute
64
TUNNELLING

• Site procedure

1. Establishing the range pole
65
TUNNELLING

• Site procedure

2. Taking the left picture
66
TUNNELLING

• Site procedure

3. Taking the right picture
67
TUNNELLING

• Processing

On site
Off site
4. Processing on the computer
68
TUNNELLING
Complete 3D acquisition including measurement
possibilities
69
RESULTS

• 3D Assessment

Determination of joint sets
Measurement of points, lengths, areas
Orientation measurements from traces and joint
surfaces
70
TUNNELLING

• Measurement of areas

57.33 m2 100
21.63 m2 37.7
71
TUNNELLING

• Geotechnical parameters

Integrated determination of rock mass parameters
Spacing, Frequency,Fisher constant Hemispherical
plots
72
TUNNELLING

• Merging of partial faces

73
TUNNELLING

• Capturing every excavation round

TM 728.5
TM 732.8
TM 739.2
Continuous analysis of discontinuity
network Gapless documentation
74
TUNNELLING
Face 03
Face 04
Face 02
Face 01
Every tunnel face Tunnel co-ordinate system
75
TUNNELLING
Computation of discontinuities on tunnel walls
Basis for determining the discontinuity system
76
TUNNELLING
Prognosis of possibly instable blocks
Selective use of additional support
77
SUMMARY
78
SUMMARY

• Data acquisition
• 2 pictures freely positioned
• calibrated digital camera
• flexible imaging distances B D/8

79
SUMMARY

• Properties
• Mobile 3D acquisition system
• Comprehensive documentation by 3D images
• Rock mass characterisation software
• Measurement of distances, areas, orientations
• Determination of joint spacing
• Hemispherical plots
• Data export
• Increase of working safety
• Increase of quality
• Reduction of acquisition costs

80
SUMMARY

• Applications
• Tunnelling
• Mining Quarrying
• Geotechnical engineeringslopes, road cuts, rock
  fall areas

81
DEMONSTRATION
Click on window for starting the program
82
Basics, principles and application of 3D imaging
systems with conventional and high-resolution
cameras
Thanks for your attentionFree demo www.3gsm.at