Complex spatial filtering with binary masks

Area of Science:

• Optics and Photonics
• Computer Vision
• Digital Image Processing

Background:

• Traditional holograms are typically produced using interference experiments.
• Holographic data storage and processing are areas of active research.
• Optical matched filtering is a key technique in pattern recognition.

Purpose of the Study:

• To investigate the feasibility of computer-generated binary holograms.
• To evaluate the image reconstruction quality of these binary holograms.
• To assess the performance of binary holograms as optical matched filters for character recognition.

Main Methods:

• Holograms were generated using a computer-guided plotter, producing binary transmittance values.
• The computer-generated plots were minified and recorded onto film.
• The binary holograms were tested in a coherent image forming system as spatial filters.

Main Results:

• Binary holograms yielded reconstructed images of quality comparable to conventional holograms.
• The binary Fourier hologram functioned effectively as an optical matched filter.
• The binary matched filter demonstrated suitability for optical character recognition tasks.

Conclusions:

• Computer-generated binary holograms provide a viable alternative to traditional holographic methods.
• Binary holograms offer comparable image quality and functionality as spatial filters.
• This technique shows promise for applications in optical character recognition and beyond.