After completing the build I went and tweaked the code to make the damn thing work.
For the processing of the image to be as fast as possible I reduce the size of the image captured to 480by320 at 60 fps.
It is fast enough to sometimes get twice the same reading as you will see in the video.
The marking of the sheets goes from one to three orange dot, it can be increased easily on the software but it need to stay on the field of view of the camera obviously.
I tried extensively the machine and it works well. The success rate increase when the area to read from is well illuminated by a white light. Here my outdoor head light 
.
So far I am really pleased with how it turned out. But the image processing side of the project is taking a lot of resources from the raspberry making it difficult to implement something else into the software...
Okay that is it for now. I will try to do some more before the deadline. But I am not confident taht I will be able to implement the direction mechanism in time.
If this is the end, I want to thank everyone that have liked, commented or just read what I wrote along the way.
Thanks Element14, Matlab and Adafruit for the opportunity.
I hope I didn't decieved anyone. And I may continue to post under the tag exercise_sheet_sorter in the future.
Time to wrap it up... Seen you on the next one.
Top Comments