Flame Image Processing using MATLAB Tools (Demo)

This tutorial describes basic image processing steps focusing on transient hydrogen-assisted flames.

This demo is provided for "Edge Detection of Hydrogen-assisted Flames: A Critical Evaluation of Several Methods to Locate Blurry, Evolving Objects" MATLAB webinar.

Part 1: Image datastore, preprocessing, intensity histograms, color spaces

Part 2: Image binarization by MATLAB built-in functions (Otsu, adaptive, k-means clustering on RGB and grayscale images)

Part 3: Image segmentation (k-means clustering in different color spaces, mulilevel Otsu thresholding)

Part 4: Reference image method - assessment of binarization techniques

Part 5: Evaluation (flame extensions, intensity histograms after background elimination, global and local statistical measures, structure extraction)

Note: Before starting the demo, copy the "Images" folder to your working folder from: https://energia.bme.hu/~kardos/MATLAB%20Flame%20Image%20Processing%20Demo%20Material/