the story is simple, I'll work for an enterprise that make a metal part, they are using ABB robots ( that use ABB rapid language ) and I'm working mostly in Python for my part, my boss asked me to find a way to speed up the robot programming because it's a pain. And I'm here to ask advice not a solution.
the actual process is (" Design > laser cut ( mcode) > grinding ( ABB rapid ) ") In my mind the idea would be to use picture recognition of the design (edge detection ) for dataset making ( we have thousands of products for that ) then "cutting" this image in part that will create the code needed for robots actions for each part, one in mcode and one in ABB rapid both with minimal data to enter by the robot dev.
my question is, would it be possible from the image recognition edge detection to create a high definition curve and from it to create a matrix, for the next part eventually making another dataset from the past script that will recognize the "type of point = type of move".
is the idea seems correct to you actually?
