My current plan is to print out a pattern and place it on the bed. I will temporarily mount a webcam on the effector and then go through a calibration routine that gathers at least 1000 data points. I will then use some optimization algorithm to solve for the best calibration values of Simpson.
I have given up on using a webcam in real time. In fact the way I am doing it requires no coordination between the firmware and the camera. It is easy enough for the webcam to see that it has been stationary for a few frames and take a datum.
Any cheap webcam should work. No mater what camera, I will need to calibrate its field of view.
I have given up on using a webcam in real time. In fact the way I am doing it requires no coordination between the firmware and the camera. It is easy enough for the webcam to see that it has been stationary for a few frames and take a datum.
Any cheap webcam should work. No mater what camera, I will need to calibrate its field of view.