Volume 2, Issue 3

Many calibration and registration methods for optical sensors require highly accurate and robust detection of markers. To date, many different approaches for detecting markers have been developed. However, all of these share certain disadvantages, depending on the optical sensor and their application. We have developed a novel approach for high-accuracy target point detection based on normed cross-correlation from symmetric patterns. The method has been specifically developed for terrestrial laser scanning, but may also work for other types of optical sensors due to its vision based properties. Our method is robust to noise and invariant to rotation, translation and perspective projection. It is scale invariant and also works in varying distances. In addition, the cross-correlation function allows quality control. Another advantage of this method is that only a few parameters need to be adjusted. Consequently, it is applicable in field test scenarios. The method does not require a certain target pattern. The only constraint on the target type is that after projecting it perspectively, a region around the target center has to remain symmetric and the approximate position of the target center has to be known. Our results demonstrate the robustness and accuracy of this method which we have validated in field applications.

The determination of uncertainties according to the “Guide to the Expression of Uncertainty in Measurement (GUM)” is extended to correlated measurements. The correlations are estimated by repeated observations. Uniform, trapezoidal and triangular distributions recommended by GUM are generalized to multivariate distributions by Monte Carlo simulations. To introduce the Type B component of uncertainty of GUM, which expresses uncertainty based on experience, two methods are discussed. The first one introduces individual corrections to the measurements with identical variances, the second one an identical correction with the same variance to each observation. Correlations are caused by the latter method. The correlations of the coordinates of points measured by the laserscanner Leica HDS 3000 are estimated from repeated observations. Except for the correlations of the three coordinates of the same points, the correlations turn out to be small. Nevertheless, they should be considered for the uncertainties of distances derived from the coordinates, especially if an addition constant for the measured distances of the laserscanner is introduced.

Establishing geodetic control networks for subsequent surveys can be a costly business, even when using GPS. Multiple stations should be occupied simultaneously and post-processed with scientific software. However, the free availability of online GPS precise point positioning (PPP) post-processing services offer the opportunity to establish a whole geodetic control network with just one dual-frequency receiver and one field crew. To test this idea, we compared coordinates from a moderate-sized (~550 km by ~440 km) geodetic network of 46 points over part of south-western Western Australia, which were processed both with the Bernese v5 scientific software and with the CSRS (Canadian Spatial Reference System) PPP free online service. After rejection of five stations where the antenna type was not recognised by CSRS, the PPP solutions agreed on average with the Bernese solutions to 3.3 mm in east, 4.8 mm in north and 11.8 mm in height. The average standard deviations of the Bernese solutions were 1.0 mm in east, 1.2 mm in north and 6.2 mm in height, whereas for CSRS they were 3.9 mm in east, 1.9 mm in north and 7.8 mm in height, reflecting the inherently lower precision of PPP. However, at the 99% confidence level, only one CSRS solution was statistically different to the Bernese solution in the north component, due to a data interruption at that site. Nevertheless, PPP can still be used to establish geodetic survey control, albeit with a slightly lower quality because of the larger standard deviations. This approach may be of particular benefit in developing countries or remote regions, where geodetic infrastructure is sparse and would not normally be established without this approach.

This paper introduces a GPS time synchronization method intended for airborne and ground-based mobile mapping systems, where multiple sensors simultaneously acquire geospatial data, and accurate co-registration of the data in space and time domains is necessary. Although more and more sensors are coming with built-in GPS capabilities, which provides for ideal GPS time-tagging, and thus for spatial registration of the data, the great majority of the sensors used in mobile mapping still lack this capability. The proposed method offers an effective and simple solution to precise GPS time-tagging of sensors that are connected to a PC-based data acquisition system. Using a timer board and a GPS receiver connection, a time base transformation is established between the high-performance processor time and the GPS time, which results in the system level availability of the GPS time for any data acquisition process.

The Bandung Basin is a large intra-montane basin surrounded by volcanic highlands, in western Java, Indonesia, inhabited by more than five million people. Based on the results of five GPS surveys conducted in February 2000, November 2001, July 2002, June 2003, and June 2005, it can be concluded that for the period of 2000 to 2005 several locations in the Bandung Basin have experienced land subsidence. Over this 5-year period land subsidence in a few locations reached a magnitude of about 70 cm, with a subsidence speed of about 1–2 cm/month. A similar rate of subsidence was also detected by the InSAR (Interferometric Synthetic Aperture Radar) technique during the period between June 2006 and March 2007. The statistical testing showed no strong correlation between GPS-derived land subsidence and registered groundwater extraction during the period of February 2000 to July 2002. The InSAR technique, however, detected significant subsidence in the textile industry area, which very large volumes of groundwater are usually extracted.

Elastic dislocation modelling of high accuracy repeated levelling data, constrained by seismological and structural data, permits a refinement of the seismic fault model of the 1986, Kalamata (SW Greece mainland), M S = 5.8 destructive earthquake. The authors' preferred solution is a blind normal fault, selected on the basis of minimum misfits between observed and predicted elevation changes. This fault pattern is broadly consistent with seismological and geologic evidence.