Template Matching on Parallel Architectures,

Abstract

Many important problems in computer vision can be characterized as template matching problems on edge images. Some examples are circle detection and line detection. Two techniques for template matching are the Hough transform and correlation. There are two algorithms for correlation: a shift and add based technique and a Fourier transform based technique. The most efficient algorithm of these three varies depending on the size of the template and the structure of the image. On different parallel architectures the choice of algorithms for a specific problem is different. This paper describes two parallel architectures: the WARP and the Butterfly and describes why and how the criterion for making the choice of algorithms differs between the two machines. (Author)

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jul 01, 1985
Accession Number
ADA170735

Entities

People

  • David Sher

Organizations

  • University of Rochester

Tags

Communities of Interest

  • C4I
  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Algorithms
  • Arithmetic
  • Complex Numbers
  • Computations
  • Computer Programming
  • Computer Science
  • Computers
  • Convolution
  • Correlation Techniques
  • Data Storage Systems
  • Fast Fourier Transforms
  • Floating Point Operations
  • Fourier Transformation
  • Image Processing
  • Lepidoptera
  • Numbers
  • Shift Registers

Fields of Study

  • Computer science

Readers

  • Approximation Theory.
  • Image Processing and Computer Vision.
  • Parallel and Distributed Computing.

Technology Areas

  • AI & ML
  • AI & ML - Machine Learning Algorithms