Regarding Can Detection

The features are very low comapred to what is shown in the notes. The can is not able to be detected.


So we recently made the coke_can a normal size and not a huge object. It wasnt realistic.
So you just have to make the Fetch Robot get closer to the table, move the headto point to the can better and move the can as close as possible to the fetch robot. For thatyou just have to modify the

def Face_Recognition_demo():
    move_base_pub.move_forwards_backwards(speed_ms=0.3,position_x=1.4, allowed_error=0.1)
    head_action.look_at(2.0, -0.2, 0.8, "map")

And then move the cocke can as explained in the notebook. You wont have as many points as with theoriginal big coke can but enough to make a detection.