Optimal cloud topographic model by statistical evaluation for large scale mapping using geodetic terrestrial laser scanner

Abstract

Terrestrial laser scanning (TLS) technology is increasingly being used for diverse types of applications such as surface reconstruction, forestry, metrology, cultural heritage preservation, reverse engineering, mine volume estimation, topographic mapping, architecture, urban planning, forensics, visualization and modelling artificial features. This technology has caused a paradigm shift in surveying from measurement of individual points to fast acquisition of accurate and highly dense 3D points. Acceptance of this technology for topographic surveying and mapping of large area and scale warrant the development of standardized specifications for data capture. Presently, most surveyors adopt the methodology of scanning as dense as possible due to fear of incomplete data, which is not the appropriate approach. Besides, the technology of geodetic TLS is almost matured, and not much research has been reported in deciding the optimal geometrical arrangements for undertaking surveys. Furthermore, previous scanning practices are qualitative in nature and present no or limited guidance or standards and practices for surveyors towards optimization. This study generated an optimal cloud topographic model using geodetic TLS technology through mathematical modelling, statistical and/or experimental evaluation for large areas and scale topographic surveying towards the development of Digital Terrain Models (DTM). In this study, scanning geometry parameters were studied/evaluated through mathematical modelling and practical experimentation in the field along with the evaluation for fast and accurate registration/georeferencing technique. Initially, during a survey, surveyors in the field can regulate scanning geometry parameters of TLS whereas other factors such as object properties, atmospheric effects and scanning mechanisms cannot be controlled. However, the critical scanning geometry parameters of TLS which include resolution, range, incident angle, laser footprint, scanner location and/or overlaps can be controlled. Experiments were carried out to verify the developed mathematical models and investigate the effects of scanning geometry parameters on the survey results. The result of these experimental investigations verified the mathematical models, which can assist surveyors prior to locating the optimal position of the scanner even for specific surveys such as archaeological sites, historical buildings and other types of survey to attain optimal results. In addition, topographic surveys through experiments and statistical analysis produced optimal range identified as ±100 m with high speed mode, optimal spatial density corresponding to angular increment of 50mm @ 10 m, maximum incident angle was ±85° and registration/georeferencing technique was occupation-backsight for large area and scale DTM. These mathematical models and experiment results can act as standards and practices guiding surveyors to carry out large area and scale topographic surveying or other specific surveys. These will help in reducing time for data collection and processing, labour and final cost of project besides assuring completeness of data. The developed mathematical models may be incorporated in the new generation of TLS that are likely to have capabilities of total station, which will further help surveyors to manage scanning geometry parameters for optimal cloud.