2017 IEEE Pacific Visualization Symposium (PacificVis) (2017)
Seoul, South Korea
April 18, 2017 to April 21, 2017
Tzu-Hsuan Wei , The Ohio State University, United States of America
Chun-Ming Chen , The Ohio State University, United States of America
Jonathan Woodring , Los Alamos National Laboratory, United States of America
HuiJie Zhang , Northeast Normal University, China
Han-Wei Shen , The Ohio State University, United States of America
Local distribution search is used in query-driven visualization for identifying salient features. Due to the high computational and storage costs, local distribution search in multi-field datasets is challenging. In this paper, we introduce two high performance, memory efficient algorithms for searching for local distributions that are characterized by marginal and joint features in multi-field datasets. They leverage bitmap indexing and local voting to efficiently extract regions that match a target distribution, by first approximating search results and refining to generate the final result. The first algorithm, merged-bin-comparison (MBC), reduces the computation of histogram dissimilarity measures by clustering bins. The second algorithm, sampled-active voxels (SAV), adopts stratified sampling to reduce the workload for searching local distributions with large spatial neighborhoods. The efficiency and efficacy of our algorithms are demonstrated in multiple experiments.
Histograms, Indexing, Data visualization, Approximation algorithms, Electronic mail, Feature extraction, Clustering algorithms
T. Wei, C. Chen, J. Woodring, HuiJie Zhang and H. Shen, "Efficient distribution-based feature search in multi-field datasets," 2017 IEEE Pacific Visualization Symposium (PacificVis)(PACIFICVIS), Seoul, South Korea, 2017, pp. 121-130.