One of the biggest differences that has been made between this year and last year is how we are handling image recognization. Last year we used opencv but this year we invested in a $60 pixy2 camera. With this camera it took us only one 3 hour session to get line following working. With the base line following that was created, we adapted that and changed the vector data that was given into bloc data. From this we were able to get the nebula challenge prototyped in only one 3 hour session. We used similar methods from the nebula challenge and were able to get a semihardcoded version of the maze done in one session.
Although we do have prototypes of all three automatic challenges done, we are hard at work improving all 3 of them. For the line following we recruited fellow student Jack (Who has helped us throughout our build) and he was able to refine our code into much more consistent results. We are currently working on incorporating ultrasonic sensors into out maze algorithm.
My name is Abigail, I'm a CNM student currently in pursuit of a biology degree. My current role in the team is to clean up/organize the lab, blog/take photos, and I'm one of the drivers for our robot.
Daily Quotes and Sayings from the team