CPU- and GPU-Based Triangular Surface Mesh Simplification

Open Access
Nistor, Dragos Mihai
Graduate Program:
Computer Science and Engineering
Master of Science
Document Type:
Master Thesis
Date of Defense:
April 02, 2012
Committee Members:
  • Suzanne Michelle Shontz, Thesis Advisor
  • John Joseph Hannan, Thesis Advisor
  • mesh
  • simplification
  • gpu
Mesh simplification and mesh compression are important processes in the realms of computer graphics and high-performance computing, as they allow the mesh to take up less memory. In particular, current simplification and compression algorithms do not take advantage of both the central processing unit (CPU) and the graphics processing unit (GPU). We propose and analyze the results of two mesh simplification algorithms based on the edge-collapse operation that take advantage of the GPU by allocating a portion of the computation to the CPU and a portion of the computation to the GPU. Our algorithms are the naive marking algorithm and the inverse-reduction algorithm. Experimental results show that when the algorithms take advantage of both the CPU and the GPU, there is a decrease in running time for simplification compared to performing all of the computation on the CPU. The marking algorithm provides higher simplification rates than the inverse-reduction algorithm, whereas the inverse-reduction algorithm has a lower running time than the marking algorithm.