EN 
Українська
PolskiXRF E-Waste Sorting: Tantalum Recovery (2D)
A 2D signal-detection model of an XRF sorting line: each particle gets a random sensor signal drawn from its true material class, and the same threshold that drives the on-screen sorting also drives the live recovery/purity numbers below.
Tantalum capacitor recycling starts with sorting: an XRF sensor scans mixed e-waste on a belt and each item gets a measured signal that overlaps between true tantalum parts and everything else. This 2D companion turns that into an explicit signal-detection model — a threshold on the sensor's noisy reading — so the same trade-off between catching more tantalum and keeping the output clean that a real sorting line faces is directly adjustable and directly visible.
Two overlapping Gaussian signal distributions (tantalum vs. other material) are classified against an adjustable threshold; recovery is the true-positive rate, purity follows from Bayes' rule using the feed grade as a prior, and every particle on the belt is sorted live using that same draw.
2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install