Authors:
Oscar Amoros
1
;
Sergio Escalera
2
and
Anna Puig
3
Affiliations:
1
Barcelona Supercomputing Center - CNS, Spain
;
2
UB-Computer Vision Center, Spain
;
3
University of Barcelona, Spain
Keyword(s):
Volume rendering, High-performance Computing and parallel rendering, Rendering hardware.
Related
Ontology
Subjects/Areas/Topics:
Computer Vision, Visualization and Computer Graphics
;
High-Performance Computing and Parallel Rendering
;
Rendering
;
Rendering Hardware
;
Volume Rendering
Abstract:
In volume visualization, the voxel visibility and materials are carried out through an interactive editing of Transfer Function. In this paper, we present a two-level GPU-based labeling method that computes in times of rendering a set of labeled structures using the Adaboost machine learning classifier. In a pre-processing step, Adaboost trains a binary classifier from a pre-labeled dataset and, in each sample, takes into account a set of features. This binary classifier is a weighted combination of weak classifiers, which can be expressed as simple decision functions estimated on a single feature values. Then, at the testing stage, each weak classifier is independently applied on the features of a set of unlabeled samples. We propose an alternative representation of these classifiers that allow a GPU-based parallelizated testing stage embedded into the visualization pipeline. The empirical results confirm the OpenCL-based classification of biomedical datasets as a tough problem wher
e an opportunity for further research emerges.
(More)