text-to-portrait

Allow an UR5 arm to draw any picture that a user wants.

Description

Text to portrait is a script that allows users to input what image they want Castor or Pollux (UR5 arms) to draw. Using the google custom search api, it downloads the first 10 images of a specific search query. It then uses opencv to extract the edges of the images and translates it to coordinates for the arm to move in. A gui is also displayed to help users select which image to draw and to control certain parameters like thresholds and drawing speed.

Demo Video

Tools Used

• Ubuntu 16.04
• ROS kinetic
• Python
• OpenCV
• Matplotlib

Team Members

• Richard Gao
• HK Rho