I need to programe a raspberry pi board with a camera to detect upside-down cans moving on a conveyor. The can is having a printed design outside cylindrical wall of the cans and there about 10 different designs. The Can can be rotated 360Deg so it can be in any orientation when facing the camera. System should be able to be teached for new print designs.
I am a programmer with extensive machine vision experience and embedded systems such as the raspberry Pi. I am also confident with machine learning tools. Please feel free to contact me to discuss this further.
Will the can orientations be the same or can they be rotated to any angle?
Hi,
I have made a similar program of object detection with Raspberry Pi and Tensorflow.
Butt it was a copy write of a company which I work.
We need to detect object using video from Raspberry Pi. and to process that we will need to have a server which will process the image/video because this operations are very heavy and Raspberrry Pi is not able to handle it speedly.
Will create a yolo based model for object detection. The model would be fast and can be retrained.
You will need to provide me curated training dataset. Further information can be discussed on pm.
Relevant Skills and Experience
I am an Embedded AI engineer working at Samsung R&D. I have lots of experience working in similar projects.