Automated Data Collection using Computer Vision
Background
This project focuses on monitoring the carbon dioxide levels in the air while 3D printers are operational in the laboratory. Data acquisition will be achieved from the following sources: sensors, air quality monitors, and a power meter. I was in charge of collecting the data from the air quality monitor. The results from the three data streams will be fed into a machine-learning model for training purposes.
Figure 1: Lab's set up of the experiment
Technical Details
-
The CO2 levels displayed on the air quality monitor were collected using computer vision.
-
A webcam was positioned in front of the air quality monitor to record CO2 levels during a 3D printing operation (Figure 1).
-
Python code was developed to extract frames from the video, capture the CO2 readings, and automatically save them to an Excel sheet. This file was then used as input data for training a machine-learning model.
Outcomes/lesson learned
-
Gained exposure to a new programming language (Python), including neural networks and the OpenCV library.
-
Performed image manipulation and processing.
-
Handled 3D printing tasks.
-
Contributed to the development of a research paper.
-
Collaborated with a professional research team, including two PhDs and a Master’s student.
-
Conducted literature reviews and provided bi-weekly research updates to the supervisor.