Drawyer (named after Draw + Sawyer, the physical Rethink Robotics robot) is a state-of-the-art automated artist. Programmed in mainly ROS and Python, our robot's goals are to:
allow a user to take a selfie
convert the face in the captured image to something Sawyer can draw
draw the user’s face on a piece of paper on a table (using a brush pen connected to Sawyer’s gripper)
Drawyer isn’t like any other drawing robots you may have seen before—Drawyer can take in an input image with some background and successfully filter out any non-facial features in the image, providing you with a simple sketch of your face.
At its core, our project is about converting a digital image to a physical medium in an unconventional way. Before the invention of cameras and printers, skilled artists would draw portraits for those people that could afford them. Our project explores the idea of utilizing robotic art to train our own skilled portrait artist, one that could work for tirelessly and cheaply. To do this, we had to tackle two key problems: (1) converting a colored, noisy image into a black-and-white one that has only facial features outlined, and (2) getting Sawyer to move its arm and draw this image in the most accurate and efficient way possible. While using a 6DOF manipulator seems like overkill, using the Sawyer platform creates a foundation for working with other mediums and canvases in 3D space.
With more advanced modifications, our robot can also be incredibly useful in other applications, such as taking security camera footage as input to draw faces of suspects and help in criminal investigations.