AUTOMATED 3D OBJECT RECONSTRUCTION VIA MULTI-IMAGE CLOSE-RANGE PHOTOGRAMMETRY

Three important stages within automated 3D object reconstruction via multi-image convergent photogrammetry are image preprocessing, interest point detection for feature-based matching and triangular mesh generation. This paper investigates approaches to each of these. The Wallis filter is initially examined as a candidate image pre-processor to enhance the performance of the FAST interest point operator. The FAST algorithm is then evaluated as a potential means to enhance the speed, robustness and accuracy of interest point detection for subsequent feature-based matching. Finally, the Poisson Surface Reconstruction algorithm for wireframe mesh generation of objects with potentially complex 3D surface geometry is evaluated. The outcomes of the investigation indicate that the Wallis filter, FAST interest operator and Poisson Surface Reconstruction algorithms present distinct benefits in the context of automated image-based object reconstruction. The reported investigation has advanced the development of an automatic procedure for high-accuracy point cloud generation in multi-image networks, where robust orientation and 3D point determination has enabled surface measurement and visualization to be implemented within a single software system.