Seeking Assistance with Image Processing for Liquid Jet in Crossflow Analysis

Hello,

I am seeking assistance with image processing techniques in MATLAB to analyze an experimental image of a liquid jet in crossflow.

Experimental Setup & Goal:The image was captured using a high-speed camera and pulsed laser illumination. The jet is contained within a rectangular chamber with glass walls on the front and back sides. My primary objective is to accurately determine the jet's breakup point and trajectory using image processing methods.

Image Challenges:The raw images present several significant issues that complicate standard analysis:

1. Noise and Saturation: The image is quite noisy, and certain regions corresponding to the spray exhibit high saturation or overexposure (appearing magenta in the image). These areas should contain usable signal for the water/spray.
2. Non-Ideal Background: Despite placing a black backdrop behind the chamber, the background is not uniformly black.
3. Reflections: A pronounced mirror reflection/ghost image of the jet and droplets is visible, which interferes with the accurate segmentation of the true jet.

Note: I have also captured a separate background image without the liquid spray present, which may be useful for subtraction methods.

Required Image Processing Steps:To achieve a clean image suitable for trajectory tracking, I need to solve the following three challenges in MATLAB:

1. Saturation Artifact Mitigation: Develop a method to remove or correct the saturated/magenta artifacts to recover the true water/spray signal within those areas.
2. Reflection Removal: Effectively eliminate the background mirror reflection/ghost image.
3. Image Segmentation: Produce a clean, segmented image that isolates only the spray and droplets without losing critical detail.

Thank you for your time and expertise.