We were able to successfully identify and color convex shapes inside the white board using the various computer vision algorithms and transformations described in the design and implementation sections. It took us a while to figure out the transformation of points since originally, the sawyer was coloring points that were off from the location of the actual shape. We successfully implemented the following:

• Approximate contour of the board
• Selection of convex shapes and image pixel coordinates
• Identify line of symmetry on the given shape and divide points into two sets of points opposite one another
• Implement a homogenous transformation from image pixels to coordinates relative to the base frame
• Continuously draw lines from top points to bottom points until shape is fully colored

All the functions are featured in the following videos of coloring a smiling face!

(We took three separate videos because the robot takes a while to reset to take a new picture of the robot. We only ran the code once during these three videos.)