Print Email Facebook Twitter A Fast General Relativity Raytracing Algorithm Title A Fast General Relativity Raytracing Algorithm: Visualising Kerr Black Holes using an Adaptive Interpolation Grid Author Verbraeck, Annemieke (TU Delft Electrical Engineering, Mathematics and Computer Science) Contributor Eisemann, Elmar (mentor) Degree granting institution Delft University of Technology Programme Computer Science | Computer Graphics Date 2017-08-30 Abstract We present a fast and accurate general relativity raytracing algorithm to visualise the distortion of space around a fast spinning Kerr black hole, as seen from a position nearby. Different parallel numerical ray integration methods were tested to solve this problem. An adaptive grid of rays, stored in efficient data structures with fast access, is used to accommodate for the difference in necessary level of precision in the image. Linear interpolation as well as spline interpolation were tested to interpolate this grid and smart pixel averaging is used as smoothing method. Despite the spline method performing better at (parts of) very low level grids, the linear interpolation method is preferred, as it is faster, more robust and supports the adaptive grid. Furthermore, a parallel CPU implementation yielded faster raytracing runtimes (factor 15) compared with a GPU implementation. The program yields high quality pictures in short runtimes (12s raytracing, < 1s interpolation, 6s pixel averaging) for an HD input image on a normal desktop computer. Subject Black HolesAdaptive GridInterpolation To reference this document use: http://resolver.tudelft.nl/uuid:8e30971d-fbd7-4d6a-a649-0f5716d89aaf Part of collection Student theses Document type master thesis Rights © 2017 Annemieke Verbraeck Files PDF Thesis_AWVerbraeck.pdf 60.1 MB Close viewer /islandora/object/uuid:8e30971d-fbd7-4d6a-a649-0f5716d89aaf/datastream/OBJ/view