Showing papers by "Norbert Pfeifer published in 2005"

Segmentation based robust interpolation - a new approach to laser data filtering

Abstract: With airborne laser scanning points are measured on the terrain surface, and on other objects as buildings and vegetation. With socalled filtering methods a classification of the points into terrain and object points is performed. In the literature two approaches – i.e. a general strategies for solving the problem – for filtering can be identified. The first work directly on the measured points and geometric criteria are used for the decision, if a point is on the ground or an object point. The methods from the second approach first segment the data and then make a classification based on segments. In this paper we present a new approach for filtering. It is a combination of both approaches, specifically exploiting their strengths. A filter method following this new approach is developed and demonstrated by examples.

3d geo-information indoors: structuring for evacuation

Abstract: Interiors of buildings are often represented as two-dimensional spaces with attributes attached to them. Examples can be found everywhere, from architectural design plans to maps showing evacuation routes in emergency cases. Most of the navigation programs use primarily 2D plans for visualization and communication. Some exceptions are historical buildings and museums, which do offer navigation through textured, detailed 3D indoor models. However, these models are not structured with respect to the functionality of the building but largely with respect to the navigation/visualisation route. Structuring the interior is not a straight forward task, as many concurrent decompositions of a building have to be accommodated (e.g., functional, according to building structure, accessibility-wise), which are very dependent on the application. Furthermore the semantic/terminology is still not unified and in many cases ambiguous. A very typical example are inner open-air courts, which are open-air but inside the building. In this paper we present a semantic model of interior spaces. The structuring aims at facilitating calculation of evacuation routes and therefore particular building characteristics related to exit possibilities such as available doors and windows are considered as leading in the followed approach. The classification concept is used to build a graph for one of the buildings in the campus area. The geometry of the building and corresponding graph are organised in a geo-database management system (geo-DBMS)

Neighborhood systems for airborne laser data

Abstract: Analysis of common neighborhood definitions for airborne laser data, triangulation or raster-based, reveals deficiencies in modeling the measured objects. Concepts that originate from 2D data structures are used for modeling complex 3D objects and for handling datasets with different point densities. Realizing these shortcomings, this paper proposes a new neighborhood system for airborne laser data. Based on laser data characteristics the proposed systems consider, among other features, point density, layered and overhanging structures, and local surface trends. Parameters for the proposed systems are derived from theoretical and practical observations. The paper demonstrates the type of neighborhood that is established by the proposed systems, and shows that artifacts that are usually created by the common neighborhoods are avoided here, and that structures within the data that are usually masked are revealed. The paper demonstrates how subsequent applications benefit from the new system. Finally, the estimation of surface normals by the proposed systems is compared to the triangulation; results show a significant improvement in the reliability and quality of the estimation.

Accuracy analysis of the Leica HDS3000 and feasibility of tunnel deformation monitoring

Abstract: Deformation monitoring with total stations, while superior in accuracy, suffers from the poor spatial resolution obtainable by this measurement method within the limited time on a tunnel construction site, and deformation can only be detected at previously marked target points. Due to its high speed in obtaining unstructured point clouds from all surfaces within the field of view terrestrial laser scanning can be ‐ at least ‐ complementary in monitoring deformation. In this study we test one of the new panoramic high precision laser scanners (Leica HDS3000) for its feasibility in tunnel deformation monitoring. Accuracy is tested in a tunnel with respect to range and angle of incidence. Not all points measured by the laser scanner can serve for deformation monitoring due to measurement artifacts at edges. We propose to use segmentation for selecting the suitable points only.

Automatic tie elements detection for laser scanner strip adjustment

Abstract: Measurements with airborne laser scanners are performed in strips, usually with multiple length strips and a few cross strips. Due to i) wrong or inaccurate calibration of the entire measurement system and due to ii) the limited accuracy of the exterior orientation determination with GPS and IMU and systematic errors of these devices, adjacent strips can show discrepancies in their overlap. For removing these discrepancies strip adjustment algorithms require quantification on these offsets at various locations within the overlapping zones. We present a general method for determining these discrepancies automatically based on segmentation of the overlap. A method to determine the accuracy of these discrepancy observations is demonstrated as well. In the examples we reconstruct mean offsets between neighbouring strips of a few centimetres, which, also show substantial variation along the strip axis. The accuracy of this discrepancy observations is in the order of 2cm. The method developed for discrepancy determination can be applied to height or full 3D strip adjustment. It can be used for approaches using the original measurements, the coordinates of the measured points, or only the offsets between surfaces.

A subdivision algorithm for smooth 3D terrain models

Abstract: Current terrain modelling algorithms are not capable of reconstructing 3D surfaces, but are restricted to so-called 2.5D surfaces: for one planimetric position only one height may exist. The objective of this paper is to extend terrain relief modelling to 3D. In a 3D terrain model overhangs and caves, cliffs and tunnels will be presented correctly. Random measurement errors, limitations in data sampling and the requirement for a smooth surface rule out a triangulation of the original measurements as the final terrain model. A new algorithm, starting from a triangular mesh in 3D and following the subdivision paradigm will be presented. It is a stepwise refinement of a polygonal mesh, in which the location of the vertices on the next level is computed from the vertices on the current level. This yields a series of triangulated terrain surfaces with increasing point density and smaller angles between adjacent triangles, converging to a smooth surface. With the proposed algorithm, the special requirements in terrain modelling, e.g. breaklines can be considered. The refinement process can be stopped as soon as a resolution suitable for a specific application is obtained. Examples of an overhang, a bridge which is modelled as part of the terrain surface and for a 2.5D terrain surface are presented. The implications of extending modelling to 3D are discussed for typical terrain model applications.

Distinguishing features from outliers in automatic Kriging-based filtering of MBES data: a comparative study

Abstract: Multi beam echo sounding is the state of the art way for surveying sea bottoms. The sea floor elevation is obtained strip wise by measuring the time it takes for sound signals, emitted simultaneously in different directions, to travel to the sea bottom and back. We compare various ways of filtering erroneous soundings from MBES data sets, all based on Kriging. This research was initiated because of the problems a classic 1D cross validation method had with distinguishing blunders from features, like pipelines. We show that part of the problems can be solved by extending the 1D method to 2D and that most problems are solved by a robust, iterative filter method.

Airborne laser scanning strip adjustment and automation of tie surface measurement

Abstract: Airborne laser scanning of the earth surface and other objects on top it yields measurements of unstructured point clouds in a strip wise manner. Often multiple length strips with a small overlap are observed, sometimes augmented by a few cross strips for validation purposes. Due to inaccurate calibration of the entire measurement system and due to the limited accuracy of direct geo-referencing (i.e., the exterior orientation determination) with GPS and IMU, including systematic errors, adjacent strips may have discrepancies in their overlap. For removing these discrepancies strip adjustment algorithms require quantification on these offsets at various locations within the overlapping zones. Different methods of strip adjustment are reviewed, followed by the presentation of a general method for determining the discrepancies automatically. This method – the core of the paper – is based on segmenting the point cloud in the overlap. In the examples, mean offsets between neighboring strips in the order of a few centimeters are reconstructed. The offsets also show substantial variation along the strip. The method developed for discrepancy determination can be applied to height or full 3D strip adjustment and for approaches using the original measurements, the coordinates of the measured points, or only the offsets between surfaces. An example of strip adjustment using discrepancy observations with the method presented and a discussion of the results conclude this paper.

Height texture of low vegetation in airborne laser scanner data and its potential for dtm correction

Abstract: The influence of low vegetation on laser scanning and its disturbing effect (a systematic positive height shift) during DTM generation is well recognized. Based on our experience in a previous study of estimating the effect of medium-height vegetation (shrubs and bushes) in rough terrain by using point-cloud based co-occurrence texture methods, we now investigate the effect of lower vegetation, such as grass, in very flat marshland conditions. When vegetation is very low, and thus the effect very small, it appears difficult to separate the systematic shift from random measurement noise. When the shift becomes larger, however, it shows significant correlation with texture measures such as slope texture and standard deviation. These measures are derived from the laser data itself and do not require any additional information.