Reconstruction of three-dimensional chemiluminescence images with a maximum entropy deconvolution algorithm

Area of Science:

• Combustion Science
• Optical Diagnostics
• Image Reconstruction

Background:

• Accurate three-dimensional (3D) visualization of flame structures is crucial for understanding combustion processes.
• Traditional methods often require multiple optical access points, limiting their applicability.

Purpose of the Study:

• To develop and validate a novel technique for generating 3D flame emission images from a single direction of optical access.
• To reconstruct the 3D flame structure using advanced imaging and algorithms.

Main Methods:

• A Cassegrain optical system with a narrow depth of field was designed to capture focused in-plane and defocused out-of-plane emission.
• The object plane was scanned through the flame by translating a mirror, acquiring a series of 2D images at various depths.
• A maximum entropy algorithm, adapted for 3D imaging, was employed to reconstruct the volumetric flame structure.

Main Results:

• The system successfully generated a family of images representing different depths within the flame.
• The maximum entropy algorithm effectively reconstructed the 3D flame structure from the acquired image series.
• Spatial resolution in the imaging direction was analyzed using laminar flames with varying offsets.

Conclusions:

• This single-direction optical access method provides a viable approach for 3D flame emission imaging.
• The developed technique enhances combustion diagnostics by enabling detailed 3D flame structure analysis.
• The use of a Cassegrain system and maximum entropy reconstruction offers improved spatial resolution and accuracy.