Hit-or-miss representation of a pattern spectrum and its optical implementation

Area of Science:

• Image analysis
• Computer vision
• Optical computing

Background:

• Pattern spectrum analysis is crucial for image recognition.
• Decomposition methods are needed for efficient computation.
• Boolean logic offers a framework for complex pattern representation.

Purpose of the Study:

• To demonstrate the decomposition of a pattern spectrum into hit-or-miss transforms.
• To express the pattern spectrum using Boolean logic functions.
• To explore the optical implementation of pattern spectrum analysis.

Main Methods:

• Decomposition of pattern spectrum into a union of hit-or-miss transforms.
• Representation of pattern spectrum using Boolean logic functions and min terms.
• Optical implementation utilizing an incoherent optical correlator with feedback.

Main Results:

• Successful decomposition of pattern spectrum into hit-or-miss transforms.
• Establishment of a Boolean-logic representation for pattern spectrum.
• Demonstration of optical implementation feasibility.

Conclusions:

• Pattern spectrum can be effectively decomposed and represented using Boolean logic.
• Hit-or-miss transforms are fundamental to this decomposition.
• Incoherent optical correlators provide a viable platform for optical pattern spectrum analysis.