Robosub Day 3 for Falcon Robotics

[falconmaster]

Here is our AUV team website
https://sites.google.com/site/falcon...csauvteam/home

here is the link to the pics so far
https://plus.google.com/photos/10513...393?banner=pwa

Here is day three report
Robosub Competition Day 3 – July 24, 2013
The day started off early for half the team, leaving the hotel at 6:30 am to reserve a practice spot. Our first run was at 9:15 am. We attempted an autonomous run, but the robot wasn’t seeing the red buoy. John ended up driving the robot around and collecting more video footage to use. The vision filtering system wasn’t able to auto correct, so more programming was needed. John and Sergio set to work after the first run to see what changes they could make to the code. After lunch the team also talked about the images the robot was seeing. It was getting too much information, which made it difficult to make decisions. Fredi and Quenan talked to some members of the professional video company that is hired to work the Robosub event. They gave three suggestions for coding to fix the problem.
In the meantime, the team gathered at 1:30 for an introduction video shoot at the top of the hill. This is the 16th year for the Robosub competition, so the theme for the event is getting a driver’s license. The video company had a white convertible mustang to use as a prop. They took a team photo and then had the team get in and around the car for more photos. They also did some video footage, and each team got to decide what action they wanted to do. Many teams went before us and they didn’t want repeat action, so we had to get creative. We ended up doing the “wave” around the car. They also were doing something with the green screen and two people in the car. Desirae and Quenan were in the car and Desirae wanted it to be like they were bad drivers, so several other team members rocked the car. It should turn out great.
Then it was back to work for the programmers. They tried one of the methods suggested for downsizing the information to the robot, which was blurring. We were waiting around for our 4 pm practice time but got called up about 20 minutes early when another team wasn’t ready. We hopped into gear and got over to the competition pool for another run. We again attempted an autonomous run but still ran into difficulties. Murphy’s Law is alive and well at Robosub! As much practicing that is done in swimming pools doesn’t really prepare us for the real competition, where more things can go wrong. We were still having trouble with the vision recognition, so John took over and did some driving but also did some quick programming fixes on the fly. A problem would come up and he would fix it, and then another problem, and another fix.
Our time was about up but we wanted to try autonomous one more time. With the threat of getting the organizers really mad and yelling at us, we kept at it one more time and did an autonomous run that went through the gate, hit the red buoy and then went above it. With our time up, we pulled the robot back in and went back to the tent to regroup. John has another idea to try tomorrow to correct the vision filtering. If it works, we should be good to go. With that thought in mind, the team plans to arrive even earlier tomorrow morning to get more practice time in the pool.
The day ended by going to the beach again and barbecuing chicken. We didn’t have much daylight left, so no one went in the water. But a group of young adults came to the park with a tightrope that they secured between two trees. Some of our students joined them and saw how far they could go on the tightrope. It was a fun, relaxing evening, and a good time was had by all.
Now the students are practicing their oral presentation and the mentors are tired and just want to rest. Much more work for tomorrow. Hopefully we’ll have plenty of good news to report.

[Dumper FTW 435]

Hey guys!
I'm an alumni from FRC team 435 "The Robodogs" and a mentor. I'm competing in robosub with the NC State team. I was really excited to see a FIRST team competing. I write the vision processing code for our team, so I would love to help you guys out if you have any questions. This is my first year on the team, so my knowledge is pretty limited, but I am our teams main vision processing programmer and I based my code off our previous work which included establishing a fairly reliable system for processing images to reduce noise. So at the very least, I might be able to point you in the right direction.

If your problem is specifically with buoys, I'm not sure how much help I will be. I beleive our buoy code was done with haar training, which you don't really have time to do right now.

If you have any questions, I'd love to give you some pointers if I can. I'm not sure what you use for vision processing (we use OpenCV), but it's possible there is some overlap if you use different libraries. Feel free to message me on here, or come by the NC State tent some time.

Good luck!
Ryan Edwards