Research On The Key Thchnologies Of Track Slab Quick Detection Based On Digitial Photogrammetry

Prefabricated track slab is bearing structure for high-speed railway, the dimension accuracywill directly affect the quality and safety. The traditional method for track slab is slow, inefficientand unable to meet the production needs; thesis attempt to carry out several key technicalresearches for industry digital photogrammetry, which will make up the lack of traditionalmethods for track slab measurement.Based on the systematic study on industry digital photogrammetry technology, some keytechnologies are deeply studied in the dissertation. Sparse bundle adjustment using LM algorithmis proposed and the effectiveness of the proposed algorithm is verified, results show thatcalculation efficiency has been greatly improved compared to the traditional method. A newmethod for the initial value calculation of a point to a surface projection based on quad tree hasbeen proposed, which made the initial value calculation effectively with high accuracy. Based ondetailed analysis on the ICP algorithm and accelerated ICP algorithm, a best fitting method from3D points to CAD model has been proposed, which made the checking computation betweenmodels and surveying data be possible. For the great needs of high speed railway track slabsdetection, a detection method based on industry digital photogrammetry has been proposed. Thendetection system based on digital photogrammetry is designed and tested. The feasibility of theproposed method is verified through field experiments.