News
Poster available
A poster about the method had been presented in the poster section of 23rd International Karstological School (2015, Postojna). It is made available to download here.
Poor mans laser scanner. A simple method in 3D cave surveying.
Abstract: Geodetic surveying of caves has great importance in speleology. Among the various surveying and graphing methods the most commonly employed technique involves projecting the wall layout and other formations of the cave onto a plane. Relating and interpreting the two-dimensional plan, profile and cross-section images that make up the spatial cave layout needs a trained eye, whereas a three-dimensional model tends to be more graphic and normally requires less knowledge to understand. For the latter, 3D modelling method, surveys are carried out on site by the aid of terrestrial laser scanners (TLS). A TLS is known to operate fast and produces a dense survey that is ideally made up of several billion points to a millimetre accuracy, potentially. However, the use of laser scanners has its limitations; the high cost of equipment being the primary reason, as it can be afforded only by the lucky few. Sensitivity, size and transportability of the instruments also restrict their use.
The widely used combination of laser rangefinder, compass and inclinometer is capable of measuring only a mere few hundred points in an hour, resulting in a much sparser and less evenly distributed point cloud than of TLS devices. For the further use and correct interpretation of this sparse point cloud it is therefore necessary to reconstruct a surface interpolating our measurements. To achieve this, we have developed a robust surface reconstruction method as well as summarized the guidelines to be considered for the better spatial distribution of survey points. Both of these areas of development require a considerable amount of test measurements, for which we measured several hundred meters of conduit in three different caves of Hungary. Surveyors of the Bányász-barlang, which is Hungary's deepest cave, use the combined method with great success, for instance. Experience has shown that this novel approach takes significantly less time than the traditional mapping and does not require extensive training. As a further benefit, data processing can be appropriately automated, which indicates great potential in terms of further reducing processing time in the future, although it is already comparable in its current state to the former method regarding efficiency.