Real-space filtering of electron images from optical autocorrelation patterns

Area of Science:

• Optics
• Electron Microscopy
• Image Processing

Background:

• Autocorrelation functions reveal repeating patterns in images.
• Optical methods can process these functions for image analysis.
• Electron microscopy images often contain complex structures requiring enhanced visualization.

Purpose of the Study:

• To demonstrate inherent image filtering capabilities within an optical autocorrelation procedure.
• To describe the optical requirements for selective image filtering based on feature frequency.
• To apply this filtering technique to analyze protein molecule projections in electron microscopy.

Main Methods:

• Utilizing a two-plate, incoherent optical procedure for recording autocorrelation functions.
• Positioning apertures over peaks in the autocorrelation plane.
• Employing a plano-convex lens to form a filtered image from light transmitted through these apertures.

Main Results:

• Demonstrated real-space filtering of electron images based on feature repetition.
• Identified optical conditions for isolating or excluding features based on autocorrelation peaks.
• Successfully applied the method to define protein molecule position and orientation.

Conclusions:

• The described optical procedure inherently filters electron images, enhancing feature analysis.
• This technique provides a method for selective image enhancement in microscopy.
• The approach is valuable for structural analysis of biological molecules like proteins.