Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

3D Shape Estimation Using Single Camera Setup and Laser Mesh Pattern

Andersson, Linnéa LU (2025) In Master's Theses in Mathematical Sciences FMAM05 20242
Mathematics (Faculty of Engineering)
Abstract
In biological observation of small flying animals, structured light methods can be used in order to make high frequency, no contact measurements in flight. Using a laser pointer that projects a 5625 point mesh pattern, as well as two high-speed cameras, three
different methods for estimating the shape of 3D objects have been developed and evaluated. The cameras were calibrated, as well as the laser pointer, which was done by treating it as a third camera and finding its camera matrix. Three algorithms were developed, one of them building upon an existing method, all which are designed to work using a single camera setup together with a laser pointer. One of the methods, named the epipolar line algorithm, was evaluated using already known... (More)
In biological observation of small flying animals, structured light methods can be used in order to make high frequency, no contact measurements in flight. Using a laser pointer that projects a 5625 point mesh pattern, as well as two high-speed cameras, three
different methods for estimating the shape of 3D objects have been developed and evaluated. The cameras were calibrated, as well as the laser pointer, which was done by treating it as a third camera and finding its camera matrix. Three algorithms were developed, one of them building upon an existing method, all which are designed to work using a single camera setup together with a laser pointer. One of the methods, named the epipolar line algorithm, was evaluated using already known projections of the pattern in 3D space. When matching points in the projected pattern with points in the known laser pattern, using unique pattern features, a maximum accuracy of 99.73\% was achieved. For the different objects examined, the estimations of position and shape were similar regardless of which method was used. There seemed to be some distortion in the estimations, as well as some mismatched points. Most promising was the epipolar line method, as it worked quickly and accurately. The highest precision was found using the image correlation and distortion method, with upwards of ten times more datapoints than the other methods. (Less)
Please use this url to cite or link to this publication:
author
Andersson, Linnéa LU
supervisor
organization
course
FMAM05 20242
year
type
H2 - Master's Degree (Two Years)
subject
publication/series
Master's Theses in Mathematical Sciences
report number
LUTFMA-3568-2025
ISSN
1404-6342
other publication id
2025:E8
language
English
id
9187497
date added to LUP
2025-09-15 11:14:28
date last changed
2025-09-15 11:14:28
@misc{9187497,
  abstract     = {{In biological observation of small flying animals, structured light methods can be used in order to make high frequency, no contact measurements in flight. Using a laser pointer that projects a 5625 point mesh pattern, as well as two high-speed cameras, three
different methods for estimating the shape of 3D objects have been developed and evaluated. The cameras were calibrated, as well as the laser pointer, which was done by treating it as a third camera and finding its camera matrix. Three algorithms were developed, one of them building upon an existing method, all which are designed to work using a single camera setup together with a laser pointer. One of the methods, named the epipolar line algorithm, was evaluated using already known projections of the pattern in 3D space. When matching points in the projected pattern with points in the known laser pattern, using unique pattern features, a maximum accuracy of 99.73\% was achieved. For the different objects examined, the estimations of position and shape were similar regardless of which method was used. There seemed to be some distortion in the estimations, as well as some mismatched points. Most promising was the epipolar line method, as it worked quickly and accurately. The highest precision was found using the image correlation and distortion method, with upwards of ten times more datapoints than the other methods.}},
  author       = {{Andersson, Linnéa}},
  issn         = {{1404-6342}},
  language     = {{eng}},
  note         = {{Student Paper}},
  series       = {{Master's Theses in Mathematical Sciences}},
  title        = {{3D Shape Estimation Using Single Camera Setup and Laser Mesh Pattern}},
  year         = {{2025}},
}