Parallel realization of single pass connected component analysis on a multi-core architecture
- 1 April 2014
- conference paper
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE) in 2014 International Conference on Communication and Signal Processing
Abstract
This paper presents a parallel algorithm for the single-pass connected component analysis of a streaming image data. The sub-images are processed simultaneously to extract the features of each connected component. Hardware implementation on a multi-core XMOS processor shows that 150 images of size 70X70 can be analysed in a second. This demonstrates the suitability of the proposed algorithm for real-time image processing.Keywords
This publication has 9 references indexed in Scilit:
- In-Place Algorithm for Connected Components LabelingJournal of Pattern Recognition Research, 2010
- FPGA implementation of a Single Pass Connected Components AlgorithmPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2008
- A Linear-Time Two-Scan Labeling AlgorithmPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2007
- A linear-time component-labeling algorithm using contour tracing techniqueComputer Vision and Image Understanding, 2004
- Handel-C implementation of classical component labelling algorithmPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2004
- An FPGA-Based Image Connected Component LabellerLecture Notes in Computer Science, 2003
- One-pass encoding of connected components in multivalued imagesPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2002
- Parallel architectures and algorithms for image component labelingIEEE Transactions on Pattern Analysis and Machine Intelligence, 1992
- Sequential Operations in Digital Picture ProcessingJournal of the ACM, 1966