Modified filter synthetic discriminant functions for improved optical correlator performance.

Area of Science:

• Optics and photonics
• Image processing
• Machine learning

Background:

• Synthetic discriminant functions (SDFs) are crucial for pattern recognition.
• Existing SDF designs often lack flexibility in preprocessing individual training images.
• Achieving equal correlation peaks in the correlator output plane remains a challenge.

Purpose of the Study:

• To propose a modified filter synthetic discriminant function (SDF) design.
• To incorporate advantageous preprocessing of training-set images using filter modulation operators.
• To develop a unified SDF design applicable to various modulation functions.

Main Methods:

• Utilizing filter modulation operators (N and M) for image preprocessing and SDF construction.
• Employing a linear combination of preprocessed images to build the SDF.
• Applying a relaxation algorithm to enforce the equal correlation peaks rule.

Main Results:

• The modified filter SDF design allows for beneficial preprocessing of training images.
• The proposed method successfully satisfies the equal correlation peaks rule.
• The filter modulation operators M and N can be defined with arbitrary functional forms.

Conclusions:

• The developed modified filter SDF design offers a generalized approach to SDF construction.
• This unified design enhances flexibility by allowing any functional form for modulation operators.
• The method provides a robust framework for pattern recognition applications requiring precise correlation control.