Complex production processes are often observed by humans in order to visually detect problems (product shape, color, surface etc.). Automating this by using an industrial image recognition technology gives great benefits in continuity and comparability of results.
Your task is to extend an existing machine demonstrator by a visual quality monitoring solution. You’d select and setup a suitable industrial camera. Next you’d investigate suitable image processing libraries to extract the relevant features from the camera images. Finally, you’d design and implement an application utilizing the image data to monitor the current production quality by counting good and bad parts.
You should have experience with image processing and programming skills in at least one common language (Java, Python, C++ etc.). Machine learning capabilities are helpful as well.