Abstract
The following thesis is a summary of work undertaken over a five year period investigating the potential of one method of remote measurement to evaluate the geometrical properties of highway surface textures. The character and configuration of surface textures contribute significantly to the safety and serviceability of highway pavements. A synopsis of current evaluation methods is considered followed by a detailed consideration of close range photogrammetry in its application to surface modelling. Close range photogrammetry is used to construct models based on the Delaunay triangulation algorithm. The outputs from these models are evaluated initially against the benchmark standard of the Mean Texture Depth (MTD) derived from the Volumetric Sand Patch (VSP).The approach addresses a significant range of surfacings at varying mesh spacings and in several contexts with results and conclusions presented at the macro and in some instances the micro scale. Images were collected within laboratory and field contexts and compared for models of surfaces at locations in Scotland, areas of mainland Europe and along the eastern seaboard of the United States of America.
Images were post processed using ImageMaster™ proprietary photogrammetric software at varying mesh spacings and Triangular Irregular Networks (TIN) generated for subsequent analysis and manipulation in other spatial and surface analysis software including ArcGIS™ and Digital Surf Mountains Map™ Premium surface analysis software. The integrity of the TINs is evaluated against TINs of a group of surfacing materials generated from photogrammetry and a ZScanner™ 800 handheld 3D laser scanner with a stated resolution of 50 microns.
The thesis does not simply present photogrammetry as an alternative to traditional volumetric methods of estimating MTD but also as a convenient method of generating an array of surface information that may be compared to functional parameters in recognised British, European and International Standards.
This study and its findings support a number of conclusions. It may reasonably be concluded that the photogrammetric method for volumetric estimation of mean texture depth and surface parameters compares well with BS EN ISO 27158. It is also demonstrated that the photogrammetric approach generates a significant array of surface information that the sand patch method does not. By definition therefore the photogrammetric method offers significantly greater insight to the nature of the surface and by extension its interaction with the tyre. The conclusions will also show that the quality of the models is sound when compared to those derived from laser scanning and may be subject to evaluation against surface parameters presented in BS EN ISO 27158 but not currently aligned with other standards within the highway sector.
Finally it is demonstrated that the photogrammetric models can be spatially adjusted to other datasets to infer mechanical processes and damage risk. This study contributes therefore to a new area of knowledge in which the measured stresses on a surface may be spatially related to the surface configuration measured remotely.
| Date of Award | Jul 2013 |
|---|---|
| Original language | English |
| Supervisor | Alan Woodside (Supervisor), David Woodward (Supervisor) & Imelda Haran (Supervisor) |
Keywords
- non-contact
- photogrammetry
- triangular irregular network
- GIS
- texture
- geometric parameters
- surface parameters
- abbott-firestone
Cite this
- Standard