How is visual information captured?