App for detecting trains in the Wedgewood-Houston neighborhood of Nashville, TN.
This project uses a CNN-based image classification model to predict if a train is currently blocking the intersections of the railroad tracks at Chestnut St. and 4th Ave.
The guts of the code live in the core module. detector.py is responsible for grabbing images from webcams pointed at each intersection and running the images through the model. Predictions, formatted as a probability of train presence, [0.0, 1.0], are published over IPC using ZeroMQ. webpublisher.py subscribes to these updates and sends them along to a web server. That web server is responsible for updating the user-facing app. eventtracker.py saves images when a train is present along with some metadata. These three components run in separate processes and are spawned from spotter.py.
If you'd like to receive notfications when a train is blocking the road, download the iOS app here.
You can read more about this project here.