Sinderberry, Mark (2007) Accuracy assessment of 3D laser scanning data utilising different registration methods. [USQ Project]
|
PDF
Sinderberry_2007.pdf Download (13MB) |
Abstract
[Abstract]: Terrestrial laser scanning is a relatively new concept for surveyors, with scanners capable of capturing large amounts of three-dimensional coordinated data quickly.
When surveying with satellites (GPS) was introduced as a new surveying tool, professional surveyors took a while to grasp the concepts, and rely on the information portrayed. Now 3D laser scanning technology has arrived, the same
confidence reservations are being revealed.
While laser scanning is used extensively overseas, Australian surveyors have not embraced the technology as quickly due to the unknown capabilities and accuracies
that can be achieved, when compared to existing equipment. Cloud models are the results of data capture utilising a technique of flooding the surveyed area with millions of coordinated points. Multiple cloud models must be stitched together to create an objects or environments outer surface. The question of how to combine all these points without losing accuracy and integrity resolves in the method of registration.
This project aims to examine the results of 3D laser scanned data, using different registration methods, to demonstrate the accuracies that can be obtained, relative to traditional survey methods. This dissertation constitutes a review of three different registration techniques available and compares the results of measurements taken from the different techniques.
The registration was conducted by placing targets mounted to features and tripods, at arbitrary locations around an established building, and surveyed in using a Trimble S6 Total Station. These same targets were then scanned from different locations to produce cloud models which were then registered. The registration methods used varied in the way they were constrained and are to be known as the:
- Cloud to Cloud Registration.
- Target Registration, and
- Georeferenced Registration.
These registration methods were then compared by way of calculating the straight line distance to the survey traverse distance records, to verify the accuracy and any
anomalies that might occur. This was a test to confirm and give confidence in this new technology.
The measurements achieved from each method produced similar results but were less than the registration error vectors, giving inconclusive results.
Statistics for this ePrint Item |
Item Type: | USQ Project |
---|---|
Refereed: | No |
Item Status: | Live Archive |
Faculty/School / Institute/Centre: | Historic - Faculty of Engineering and Surveying - Department of Surveying and Land Information (Up to 30 Jun 2013) |
Date Deposited: | 29 Feb 2008 05:29 |
Last Modified: | 02 Jul 2013 22:59 |
Uncontrolled Keywords: | 3D laser scanning; cloud to cloud registration; target registration; georeferenced registration |
Fields of Research (2008): | 09 Engineering > 0909 Geomatic Engineering > 090906 Surveying (incl. Hydrographic Surveying) |
Fields of Research (2020): | 40 ENGINEERING > 4013 Geomatic engineering > 401306 Surveying (incl. hydrographic surveying) |
URI: | https://sear.unisq.edu.au/id/eprint/3947 |
Actions (login required)
Archive Repository Staff Only |