Enhancement of image recognition in digital close-range photogrammetry using mathematical equations and camera calibration
The use of close-range photogrammetry technique via mathematical modelling and camera callibration are becoming an interesting area in reverse engineering to describe relationships between spatial objects in various applications. It has been used as an alternative solution to conventional meth...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English English English |
| Published: |
2020
|
| Subjects: | |
| Online Access: | http://eprints.uthm.edu.my/1038/2/24p%20MUHAMMAD%20ADRIAN%20THOMAS.pdf http://eprints.uthm.edu.my/1038/1/MUHAMMAD%20ADRIAN%20THOMAS%20COPYRIGHT%20DECLARATION.pdf http://eprints.uthm.edu.my/1038/3/MUHAMMAD%20ADRIAN%20THOMAS%20WATERMARK.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The use of close-range photogrammetry technique via mathematical modelling and
camera callibration are becoming an interesting area in reverse engineering to
describe relationships between spatial objects in various applications. It has been
used as an alternative solution to conventional methods for generation of virtual 3D
models, which for use in CAD, CAM and CAE studies. However, many studies have
revealed that most of the developed algorithms for 3D modelling possess several
major challenges due to its tedious task during the image recognition phase. Thus,
this research focuses to investigate the model reconstruction process for an enhanced
image capturing method using close-range photogrammetry. Mathematical equations
are presented to potentially help for an effective modelling operation as they are used
to define the appropriate parameters for shooting phase based on the object’s size.
This is to obtain as sufficient number of images as possibile to increase the overall
overlapping areas between images during the photogrammetric data acquisition
period. This research used different types of camera sensors and calibrations to
capture the images respect to the reconstructed 3D models are presented. Finally, this
research proposed a thorough framework incorporated in data acquisition factors and
mathematics equations to capture a set of images using smartphone camera for close�range photogrammetry. The feasibility of the framework for generating a 3D model
using photogrammetry was verified with a 3D model generated using ATOS 3D
Scanner (high resolution 3D scanner for reverse engineering). Results obtained
showed that the photogrammetric model reaches accuracy of 99% with the model
scanned using ATOS along with maximum and minimum errors of 0.0003 m and
0.0004 m respectively. With that, using the same photogrammetric settings three case
studies of different object sizes were investigated. At the end of this study, the
objective of an enhanced image capturing method is successfully achieved. |
|---|