Lianping Xing , The Chinese University of Hong Kong, Hong Kong
Xiaoting Zhang , The Chinese University of Hong Kong, Hong Kong
Charlie C.L. Wang , The Chinese University of Hong Kong, Hong Kong
Kin-Chuen Hui , The Chinese University of Hong Kong, Hong Kong
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/MCG.2013.84
This article presents highly parallel algorithms for remeshing polygonal models guided by cues from human visual perception. The remeshing framework is based on meshfree techniques for processing surface sample points. The benefit is twofold: it is robust to input models with problematic connectivity and the geometric processing on points is easier to run in parallel on GPUs. The visual perception information is extracted in the image space and then mapped back to the Euclidean space. Based on these cues, a saliency field is generated to re-sample the input model. Lastly, a new projection operator is developed to further optimize the distribution of re-sampled points. As the number of vertices on the resultant model is controlled by the downsampled points, this remeshing framework can also be used in model simplification. Experimental results demonstrate that our algorithm can remesh diverse polygonal models to well-shaped triangular meshes with high visual fidelity.
Lianping Xing, Xiaoting Zhang, Charlie C.L. Wang, Kin-Chuen Hui, "Highly Parallel Algorithms for Visual Perception Guided Surface Remeshing", IEEE Computer Graphics and Applications, , no. 1, pp. 1, PrePrints PrePrints, doi:10.1109/MCG.2013.84