Detecting Corona Hotspots
Customer: St. Peter Ording
Role: ML Engineer/Data Scientist
COVID19 has caused a major impact on tourism. St. Peter Ording is a popular German seaside spa and has been looking to detecting the so called “Corona Hotspots” in order to control people flows on the beach. Traditional IoT sensors for counting people are working great at tight places but cannot work on such large areas as beach.
To overcome this challenge, I developed a computer vision pipeline that uses cameras instead of sensors. It receives images from the cameras, processes them, counts people and sends the results back to the dashboard. The pipeline was deployed on Microsoft Azure using Azure Functions and Azure Storage. The underlying computer vision algorithm was based on PyTorch.
