Segmentation and determination of grid points of curve points in terrestrial laser scanning data for regular curve surfaces via C-means integrated fuzzy logic approach

Abstract

Terrestrial Laser Scanners (TLS) are used frequently in three dimensional documentation studies and present an alternative method for three dimensional modeling without any deformation of scale. In this study, point cloud data segmentation is used for photogrammetrical image data production from laser scanner data. The segmentation studies suggest several methods for automation of curve surface determination for digital terrain modeling.In this study, fuzzy logic approach has been used for the automatic segmentation of the regular curve surfaces which differ in their depths to the instrument. This type of shapes has been usually observed in the dome surfaces for close range architectural documentation. The model of C-means integrated fuzzy logic approach has been developed with MatLAB 7.0 software. Gauss2mf membership functions algorithm has been tested with original data set. These results were used in photogrammetric 3D modeling process.As the result of the study, testing the results of point cloud data set has been discussed and interpreted with all of its advantages and disadvantages in Section 5.