#127:
P. Bellens,
K. Palaniappan,
R. M. Badia,
G. Seetharaman, and
J. Labarta
Lecture Notes in Computer Science (ACIVS),
Volume 6915,
pgs. 586--598,
2011
Abstract,
Bibtex,
PlainText,
PDF,
URL,
DOI,
Google Scholar
The integral histogram is a recently proposed preprocessing technique to compute histograms of arbitrary rectangular gridded (i.e. image or volume) regions in constant time. We formulate a general parallel version of the the integral histogram and analyse its implementation in Star Superscalar (StarSs). StarSs provides a uniform programming and runtime environment and facilitates the development of portable code for heterogeneous parallel architectures. In particular, we discuss the implementation for the multi-core IBM Cell Broadband Engine (Cell/B.E.) and provide extensive performance measurements and tradeoffs using two different scan orders or histogram propagation methods. For 640×480 images, a tile or block size of 28×28 and 16 histogram bins the parallel algorithm is able to reach greater than real-time performance of more than 200 frames per second.
@article{Bellens:ACIVS-2011,
author = "P. Bellens and K. Palaniappan and R. M. Badia and G. Seetharaman and J. Labarta",
title = "Parallel implementation of the integral histogram",
year = 2011,
journal = "Lecture Notes in Computer Science (ACIVS)",
volume = 6915,
pages = "586--598",
keywords = "parallelization, gpu, tracking, fmv, motion, features, dod",
doi = "10.1007/978-3-642-23687-7_53",
url = "http://link.springer.com/chapter/10.1007/978-3-642-23687-7_53"
}
P. Bellens, K. Palaniappan, R. M. Badia, G. Seetharaman, and J. Labarta. Parallel implementation of the integral histogram. Lecture Notes in Computer Science (ACIVS), volume 6915, pages 586--598, 2011.
