Started to code the pipeline that preprocesses frames from greyscale to kill pixel noise, uses OpenCV’s MOG2 to build a model of an empty chamber that lights up anything that moves inside it at an initial fps of 60 frames for warm-up, Feature extraction to classify Alpha-Beta-Muons apart on observation, and finally a logger that writes every real detection that isn’t noise. Before I test this with a Pi camera, I will be trying it with my regular PC camera.