I am in a master degree research and I'm in need of a data set for image processing of solar panels with defects (mainly hot spots), so I need to know if someone can help me and tell me where I can find one.
Does it have to be a real data set from an experimental arrangement or can you just simulate one? The advantage of a simulated one is that you can make the hot spots as big or small as you want and the temperature as far above ambient as you want. This will help clarify how good your image processor is at finding points of failure.
Thank you, Tony Maine . Do you indicate any simulator for this purpose?
I am sure you could write one! It won't take long. Set up an array to represent all the solar panels in an installation and set up an array to simulate one panel with each cell represented by say an array of 16 (4 x 4) zones. You can then duplicate this for all the panels. Then you can assign an overheating value to each zone in each cell in each panel. It won't be necessary to simulate the cell's electrical behaviour if you're just using IR to detect failures. Assume that in a real array a cell is at high risk of failure once it's temperature passes about 100C, so your IR scanner must pick out such hot spots well before this point is reached. Best wishes with your research!
- Badges
- Science topic
- Similar topics
- Computer Science
- Images