A Python-based package for the acquisition, digital signal processing, plotting, and storage of sensor data in realtime. Please see fastplotlib, developed by Kushal Kolar, to learn more about the plotting library we primarily use.
1.) Custom Serial Acquisition - Users can write and pass their own acquisition handler into the sensor_core pipeline. Refer to the custom serial handler notebook.
2.) Digital Signal Processing Integration - sensor_core has a DSPManager hook enabling custom or predefined DSP algorithms to be applied prior to visualization without affecting the underlying datastream. Refer to the DSP notebok.
3.) High-Speed Visualization - using fastplotlib, we can reliably visualize 2- and 3-D data at high speed. We have thus far tested only in Jupyter Notebooks. Refer to the line and image notebooks for visualization examples.
4.) High-Bandwidth Storage - sensor_core creates temporary .bin files to stream data rapidly to before offloading to a sqlite file, enabling stable long term storage while imposing minimal delay in the real-time processing pipeline. Refer to the line and image notebooks for storage examples.
pip install sensor_pipeline
select/cd into directory you want to install sensor_core
git clone https://github.com/BailabUNC/sensor_core
cd sensor_core/
pip install -r requirements.txt
pip install -e .
The following data was captured by MABOS: a proprietary biosensor we developed.