Optical Wavelet Preprocessing.

Abstract

A low resolution image was used for object recognition using an optical correlation technique. The low resolution image was an approximation of the original image that was derived using a discrete wavelet transform. The low resolution image was cross correlated with an image containing an object of interest. Using an inverse wavelet transform, the cross correlation result was synthesized to the resolution of the original image. This approach avoids cross correlating an image multiple times and passing information between levels of an image representation. It was shown that objects could be recognized that were approximated at 1/4 the original resolution using binary phase only correlation filters. Discrimination decreased at the lower resolutions, but the technique looks useful for object recognition.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jul 01, 1996
Accession Number
ADA313827

Entities

People

  • Samuel P. Kozaitis

Organizations

  • Florida Institute of Technology

Tags

Communities of Interest

  • Advanced Electronics
  • Energy and Power Technologies
  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Command And Control
  • Computational Science
  • Computer Science
  • Computer Vision
  • Correlation Techniques
  • Cross Correlation
  • Discrimination
  • Filters
  • Frequency Domain
  • High Pass Filters
  • Low Resolution
  • Object Recognition
  • Optical Correlators
  • Pattern Recognition
  • Recognition
  • Signal Processing
  • Wavelet Transforms

Readers

  • Computer Vision.
  • Electromagnetic Wave Scattering and Antenna Radiation Engineering