Conference Proceeding

A data parallel implementation of an edge point chaining: towards a new principle of edge linking

IUT de Villetaneuse, Univ. de Paris-Nord, Villetaneuse
10/1996; DOI:10.1109/ICIP.1996.560622 ISBN: 0-7803-3259-8 pp.141 - 144 vol.2 In proceeding of: Image Processing, 1996. Proceedings., International Conference on, Volume: 1
Source: IEEE Xplore

ABSTRACT The paper focuses on the problem of the implementation under the
data parallel programming model of an edge point chaining algorithm.
This implementation is not a straightforward transposition of classical
algorithms developed so far; indeed all of those are based upon a video
scanning of the image, and are thus sequential by nature. Therefore, a
new data parallel algorithm has been designed. The principle of the data
parallel implementation is detailed. The implementation technique is
analogous to the parallel region growing algorithm

0 0
 · 
0 Bookmarks
 · 
12 Views
  • Source
    Article: A Fast Table-Driven Method for Edge Thinning and Linking
    David Prewer, Les Kitchen
    [show abstract] [hide abstract]
    ABSTRACT: This paper describes fast but effective methods for edge thinning and linking. It has its genesis partly in the idea that it is often preferable to perform simple actions quickly rather than to attempt to do more sophisticated things that take longer. This idea is particularly attractive in the area of real-time computer vision, where the flow of input data may be otherwise overwhelming. We describe computationally cheap table-driven versions of the conventional techniques of edge thinning and edge linking that involve no transcendental functions and only integer arithmetic. Details are given of their computational speed on some representative images. 1. Introduction Edges are important features for object recognition and for matching for stereo and motion analysis. Edge features can be obtained from raw edge detector responses by a process of edge thinning and edge linking. Here we present fast table-driven techniques to perform edge thinning and linking. The basic approach of edg...
    02/2000;
Data provided are for informational purposes only. Although carefully collected, accuracy cannot be guaranteed. The impact factor represents a rough estimation of the journal's impact factor and does not reflect the actual current impact factor. Publisher conditions are provided by RoMEO. Differing provisions from the publisher's actual policy or licence agreement may be applicable.

Keywords

algorithm
 
data parallel programming model
 
edge point chaining algorithm
 
implementation technique
 
new data parallel algorithm
 
sequential