Èñòî÷íèê: http://k154.fsv.cvut.cz/~koska/publikace/soubory/Zurich_AJ.pdf
Measuring deformations is a significant part of geodetic works in civil engineering. The article assesses possibilities of using laser scanning technologies for measuring deformations. Experiments verifying accuracy of determination of displacements of HDS targets, point clouds and modelled plains were conducted. Experiments were conducted both in laboratory conditions with using interferometer and in practical conditions of gallery and tunnel. The results were assessed graphically with using hypsometric maps and on basis of calculation of empirically determined standard deviations of displacements.
Measuring deformations is an important part of geodetic works during carrying out and supervising buildings. Most used methods currently used for measuring deformations are methods using the total stations and GPS technologies. An interesting possibility in the area of measuring deformations is the laser scanning technology. This technology does not reach such accuracy for the individual points as in case of the exact total stations or of the long-term GPS observation but it overruns this drawback in point density and complexity of surface record. That is why possibilities of using laser scanning in the area up to now stipulated especially to the exact total stations are described in our paper. Our attention is concretely given to the possibility of setting this technology for measuring deformations in subterranean buildings.
The experiment took place in underground laboratory furnished with the Renishaw ML10 Gold Standard interferometer with standard deviation 0.7 m. HDS targets, a sphere and a plane were placed on the interferometer wagon. Displacements in longitudinal and transversal direction were being determined.
Three experiments were conducted on the whole. In the first experiment, the distance of the inferometer wagon from the scanner was approximately sixteen metres. Two HDS square-shaped targets 3" x 3", one circular target with diameter 6" and an exact sphere with diameter 218 mm were placed on the wagon. The wagon with targets was measured in ten positions. Displacements between the single positions were in range of centimetres in longitudinal direction. The targets were measured in all positions as HDS targets with using internal process of determining position of identical points with the Cyclone Scan software. The sphere was measured in the same way. In the second experiment, a similar measurement on the distance of approximately five meters was taken. In the third experiment, a plane was placed on the wagon so that normal of this plane were approximately parallel to direction of scanning.
One experiment was conducted on distance approximately five metres by reason of limited space of the laboratory. Two HDS square-shaped targets 3" x 3, one circular target with diameter 6" and an exact sphere with diameter 218 mm were placed on the wagon. The measurement was taken in ten positions of the wagon with mutual transversal displacements in range of centimetres.
3D positions of all measured HDS targets in all stages were subtracted in the Cyclone software. Displacements of targets between the single stages were calculated from these positions. These displacements were compared with the exact displacements gained from the measurement of the interferometer. With respect to accuracy of the inferometer it is possible to take the gained differences for real errors. Standard deviations of target displacements were calculated from these errors - see table 1. When determining circular target, an average of standard circular HDS targets (average six inches) was used in the Cyclone Scan software.
Figure 1: Photo and point cloud of interferometer wagon with targets
When determining plane displacements in longitudinal direction, its point clouds in the individual stages were inset with a plane in Cyclone. Distance of one point of one plane from the other plane was calculated so as to assess the displacement.
Table 1: Standard deviation of displacement
| Target / Measurement | Longitudinal 16 m[mm] | Longitudinal 5 m[mm] | Transversal 5m[mm] |
| 1. square target 3x3” | 0.75 | 0.82 | 0.11 |
| 2. square target 3x3” | 0.70 | 0.41 | 0.21 |
| Circular target 6” | 0.69 | 0.73 | 0.11 |
| Average from targets | 0.71 | 0.68 | 0.15 |
| Sphere | 0.74 | 0.58 | 0.75 |
| Plane (modeled) | 0.36 |
Total results from determination of displacement of plane HDS targets, a sphere and a plane can be seen in table 1. The stated data show that accuracy of measurement of a square and a circular plane HDS target is similar.
Standard deviation of displacement of a plane HDS target and a sphere in longitudinal direction is approximately 0.7 mm in distance both 5 and 16 metres. Standard deviation for a plane in distance 5 meters is 0.36 mm.
Standard deviation of displacement of one point of a cloud in longitudinal direction amounting to 3.3 mm was then found out. Standard deviation of displacement of a plane HDS target in transversal direction is approximately 0.15 mm and for a sphere 0.85 mm. Higher accuracy in transversal direction is caused by higher angle accuracy of the HDS 3000 scanner in comparison with longitudinal accuracy (see [1]).
Method of average point displacement was designed so as to evaluate normal displacements of any surface. It is based on difference DTM of the observed area. The result is not burdened with measurement noise. The method was used for evaluation of longitudinal displacements of a plane and the resulting standard deviation is 0.32 mm.