Development of a Facial Recognition Pantograph Drawing Robot

Abstract

This paper describes the design and application of a solar powered drawing robot that implements face detection, image to sketch conversion, and wireless communication between the robot and the main computer. The robot has the capability to detect and capture a picture of a human face, convert the image to a sketch in the form of a series of lines, and physically draw the image on a whiteboard. Additionally, the robot can draw any selected image stored on the main computer, apply or remove shading from the image, move to a specified location away from the sketch to provide a clear area for the user to erase the board, and can draw diagnostic test rectangles to analyze the mechanical performance of the system. These robotic functions were implemented by first determining the angles and geometry of the arms required to locate a given x-y coordinatepoint. This was done using properties of triangles, as well as the known physical limitations of the pantograph's size. A CAD model of the design was created to incorporate both the calculated geometry and purchased parts, such as the servo motors, axles, and bearings. Existing open-source libraries were used to aid in facial recognition and the conversion of images to a sketch of x-y points. By utilizing the known geometry of the robot, the pen was programmed to move to specified points based on the angles of the two servo motors.