In this master thesis, we discuss the currently available tools for automatic line detection in point clouds obtained from terrestrial laser scanning. The rapid development of data acquisition technologies and their increasing use in surveying, construction and other industries creates a need for still accurate but faster data processing. In this thesis we have examined and compared some of the commercial tools currently available on the market: Civil 3D, ReCap, UNDET and CloudCompare. In order to carry out the testing and analysis, two reference drawings, Bridge and Building, were prepared and used as a basis for the comparisons. The test cases are representative samples of the interior and exterior of an engineering building, in order to study the impact of different conditions on the performance of the algorithms. The results showed that the tools differed in terms of their performance in line detection in two and three dimensions, as well as in terms of their accuracy, reliability, speed and usability. None of the tools proved to be universally applicable to both test cases, indicating the limitations of each approach. In addition to the technical aspects, the study also included a cost analysis of the tools. Based on the findings, the tools were evaluated, highlighting their strengths and weaknesses according to the specific needs and conditions of use.
|
