TY - JOUR
T1 - Sample-based surface coloring
AU - Bürger, Kai
AU - Krüger, Jens
AU - Westermann, Rüdiger
N1 - Funding Information:
The authors wish to thank Leif Kobbelt and Peter Schroeder for providing some of the vector fields on surfaces shown in this paper. This work was funded in part by the NIH/NCRR Center for Integrative Biomedical Computing, P41-RR12553-10.
PY - 2010
Y1 - 2010
N2 - In this paper, we present a sample-based approach for surface coloring, which is independent of the original surface resolution and representation. To achieve this, we introduce the Orthogonal Fragment Buffer (OFB)-an extension of the Layered Depth Cube-as a high-resolution view-independent surface representation. The OFB is a data structure that stores surface samples at a nearly uniform distribution over the surface, and it is specifically designed to support efficient random read/write access to these samples. The data access operations have a complexity that is logarithmic in the depth complexity of the surface. Thus, compared to data access operations in tree data structures like octrees, data-dependent memory access patterns are greatly reduced. Due to the particular sampling strategy that is employed to generate an OFB, it also maintains sample coherence, and thus, exhibits very good spatial access locality. Therefore, OFB-based surface coloring performs significantly faster than sample-based approaches using tree structures. In addition, since in an OFB, the surface samples are internally stored in uniform 2D grids, OFB-based surface coloring can efficiently be realized on the GPU to enable interactive coloring of high-resolution surfaces. On the OFB, we introduce novel algorithms for color painting using volumetric and surface-aligned brushes, and we present new approaches for particle-based color advection along surfaces in real time. Due to the intermediate surface representation we choose, our method can be used to color polygonal surfaces as well as any other type of surface that can be sampled.
AB - In this paper, we present a sample-based approach for surface coloring, which is independent of the original surface resolution and representation. To achieve this, we introduce the Orthogonal Fragment Buffer (OFB)-an extension of the Layered Depth Cube-as a high-resolution view-independent surface representation. The OFB is a data structure that stores surface samples at a nearly uniform distribution over the surface, and it is specifically designed to support efficient random read/write access to these samples. The data access operations have a complexity that is logarithmic in the depth complexity of the surface. Thus, compared to data access operations in tree data structures like octrees, data-dependent memory access patterns are greatly reduced. Due to the particular sampling strategy that is employed to generate an OFB, it also maintains sample coherence, and thus, exhibits very good spatial access locality. Therefore, OFB-based surface coloring performs significantly faster than sample-based approaches using tree structures. In addition, since in an OFB, the surface samples are internally stored in uniform 2D grids, OFB-based surface coloring can efficiently be realized on the GPU to enable interactive coloring of high-resolution surfaces. On the OFB, we introduce novel algorithms for color painting using volumetric and surface-aligned brushes, and we present new approaches for particle-based color advection along surfaces in real time. Due to the intermediate surface representation we choose, our method can be used to color polygonal surfaces as well as any other type of surface that can be sampled.
KW - Sample-based graphics
KW - graphics data structures
KW - surface coloring
KW - surface particles
UR - http://www.scopus.com/inward/record.url?scp=77954589391&partnerID=8YFLogxK
U2 - 10.1109/TVCG.2009.107
DO - 10.1109/TVCG.2009.107
M3 - Article
C2 - 20616392
AN - SCOPUS:77954589391
SN - 1077-2626
VL - 16
SP - 763
EP - 776
JO - IEEE Transactions on Visualization and Computer Graphics
JF - IEEE Transactions on Visualization and Computer Graphics
IS - 5
M1 - 5262943
ER -