NEDT Characterization Bench
Automated noise-equivalent differential temperature testing for infrared cameras — capture, statistics, and reporting in one instrument.
NEDT — noise-equivalent differential temperature — is the figure of merit that says how small a temperature difference an infrared camera can actually resolve before it drowns in its own noise. Measuring it well means controlled scenes, lots of frames, and careful statistics.
This bench automates the whole procedure: synchronized frame capture, per-pixel temporal noise statistics, responsivity measurement, and the NEDT computation itself, with structured reports at the end. It has been through two major versions (V2 and V3), each tightening the acquisition flow and the analysis.
It’s a good example of the kind of instrument I like building: the physics definition is one line, and everything else is engineering discipline — making sure integration times, frame counts, and reference conditions are captured with the data so every number is reproducible later.
Python, PyQt6, pyqtgraph, NumPy. Source is private.