mot-monitor

Real-time MOT (magneto-optical trap) fluorescence monitor using USB cameras.

Turn any USB camera into a quantitative MOT diagnostic tool — no photodiode required.

Features

• Real-time fluorescence intensity — ROI-based intensity with automatic background subtraction
• Loading curve fitting — N(t) = N_ss (1 - exp(-t/τ)) with steady-state atom number and lifetime
• MOT position tracking — centroid and 2D Gaussian size of the fluorescence spot
• Background ring subtraction — robust against ambient light variations
• Live strip chart — time series of fluorescence intensity and position
• Recording & export — save time series to CSV or HDF5
• PyQt5 GUI — live camera view with ROI overlay, plots, and one-click fitting
• CLI — headless monitoring and batch fitting

Installation

pip install mot-monitor            # core (OpenCV + NumPy + SciPy)
pip install mot-monitor[gui]       # + PyQt5 GUI
pip install mot-monitor[export]    # + HDF5 export

Quick start

GUI

mot-monitor gui

CLI

# Live monitor (prints fluorescence intensity to stdout)
mot-monitor monitor --camera 0 --roi 200,150,100,100

# Fit loading curve from recorded CSV
mot-monitor fit recording.csv

# Record fluorescence time series
mot-monitor record --camera 0 --duration 30 --output data.csv

How it works

1. Capture frames from USB camera via OpenCV
2. Define ROI around the MOT fluorescence spot
3. Subtract background using an annular ring around the ROI
4. Integrate pixel intensity within ROI → fluorescence signal (proportional to atom number)
5. Track signal over time → loading curve, optimization feedback

Background subtraction

The background ring method estimates the local background from an annular region surrounding the ROI. This is subtracted per-frame, making the measurement robust against:

• Room lighting changes
• Scattered laser light
• Camera gain drift

Why not a photodiode?

MOT loading timescales are seconds — USB camera frame rates (30 fps) are more than sufficient.
A camera also gives you spatial information (position, size, shape) that a photodiode cannot.

References

• Tranter et al., "Multiparameter optimisation of a magneto-optical trap using deep learning", Nature Communications 9, 4360 (2018) — USB camera fluorescence for MOT optimization
• Chin et al., "Direct measurement of loading and loss rates in a magneto-optical trap", PRA 76, 013402 (2007)

License

MIT