Team 254 Presents: CheesyVision

[Jay Meldrum]

Originally Posted by Adam Freeman

Team 254, thank you!

We ran CheesyVision any time we were doing a 1 ball auto at MSC and it worked perfectly.

You guys are awesome.

I second this! Thanks again 254!

Originally Posted by DjScribbles

I haven't seen any talk of a C++ port, so I started a thread in the C++ sub forum here to avoid derailing this thread:

http://www.chiefdelphi.com/forums/sh...26#post1372026

My prototype code is linked in the thread, it is completely untested, but any contributions are welcome.

Thanks Poofs, very awesome implementation; looking forward to trying this out.

Also thank you to DjScribbles for base of our code in C++. Just had to change a few things and it worked great.

Thanks!

[Mike Copioli]

We added it this weekend to our Prac bot. Looks like we will have two ball hot goal detect at Champs.

Thanks so much guys.

[IKE]

I heard a lot of teams at MSC were using CheesyVision this weekend with great success. Kudos to releasing such a nice product in season.

Class Act.

CheesyVision is much better than Teh CheeZViSHUN (which apparently just tinted all camera inputs blue).

[DjScribbles]

As a heads up, I found that our netbook's webcam would cause an exception when the cheesy vision script started, to resolve this, I added a delay between initiating the camera connection, and grabbing the first image.

I don't have the script handy to share, but as someone who's never used python before, I'm confident that someone else could improve on the implementation anyway.

I also saw some weird connection issues between the Client and Server on Saturday on the field (all other systems were normal, just lack of cheesy-vision). I left more details on the issue here.

[connor.worley]

I learned to write software for FRC bots by reading 254's 2010 code. It's great that releases are still coming out each year.

[Andrew Schreiber]

Originally Posted by TravSatEE

Your analogy to a COTS part is not equivalent to this situation: several mentors appear to have worked exclusively on a project that was used to give a competitive advantage to the game performance given limitations of the Field Management System.

Uh, what?

You were around in the days back before the off the shelf shifters. You should recall that before AM started selling their shifters there were few teams that could reliably shift. This is EXACTLY the scenario. Only instead of only rich teams having access to it (a set of shifters will run you what, $700?) anyone with an internet can use this.

Comments like this make me question whether I should open source anything lest I be accused of 'cheating'.

[JamesTerm]

Originally Posted by Andrew Schreiber

Comments like this make me question whether I should open source anything lest I be accused of 'cheating'.

Please don't let this one person's comment question or prevent you from posting code... look at all the good that has come from Jared releasing this code.

If we give in to comments like this... and good code stops being shared everyone will lose... Keep the good code flowing! Exposure to continuously well-written code for students to view can inspire them. Now that I think about it, well-written code still inspires me!

[billbo911]

Originally Posted by JamesTerm

... Exposure to continuously well-written code for students to view can inspire them. Now that I think about it, well-written code still inspires me!

I could not have said it any better!

[Dunngeon]

Originally Posted by TravSatEE

I had expected that your project was forked and that was why I asked for clarification as to what the students did. Instead, your answer wasn't completely clear to me as to exactly what the students did for CheesyVision. I do understand that it was "mentor heavy." Though you couldn't tell the differences between student and mentor effort when you were in high school, I trust my judgment because I have done programming for 18 years and know the subtle differences in programming skills at all levels. I do think very highly of the work you released to all teams. I am sure students also do.

I don't understand why the Poofs have to prove to you that student's built it. I'm pretty sure your team used gearboxes from Vex or AndyMark this year, which were designed by mentors/paid engineers. Should those also be exclusively designed and built by students? The idea of COTS is to allow easier entry into FRC and raise the level of play. I would argue that it has, because the level of play is exponentially higher than in the early 2000's when teams were required to build (almost) everything. The same concept applies here, a team that built something amazing is sharing it with the greater community in an effort to increase the competition level. Beyond that, they could have waited until the end of the season to release this vision program, keeping a competitive edge over most teams. Instead they have released it, and had mentors comb through it so that it is easy to implement. I think this release speaks volumes about the character of the members of Team 254.

Your analogy to a COTS part is not equivalent to this situation: several mentors appear to have worked exclusively on a project that was used to give a competitive advantage to the game performance given limitations of the Field Management System. Albeit it was not an overwhelming advantage and any team could have done the exact same thing. Again, I am not saying Team 254 has broken any rules. But I find it interesting that a NASA sponsored (funded?) team, and the team with the best winning record of FIRST, needs to have mentors do exactly what you have done for a high school competition. Of course you stand by your decision to do CheesyVision the way that you did -- it's easy to stand by a decision that has no consequences.

Again, you wouldn't of known this even existed if they hadn't released it.

I am eminently fortunate to always have mentored teams that were student run and each team has students just as impressive as the ones you described. From what I have learned today, I think the difference between your team and my teams is that other mentors keep it students vs students.

Either team type can have benefits and drawbacks, it's all in implementation. Our team is fully student run, but sometimes I wish we had more mentors because then other students and I could learn so much more

I do not intend for any of my posts to put you on the defensive nor to diminish your students' work hard. I am trained to speak my mind and your reply has been informative. Thank you for answering.

I'm gonna speak my mind here, everything you wrote above this diminishes the work students have put into Team 254's robot this year. I've no doubt that much of the robot was mentor driven, either directly or indirectly, but if mentors built the entire robot I greatly doubt that top students would stay around for long. The students of 254 make it what it is, just like the students of 955 make our team what it is. Mentors add capabilities to teams because of the knowledge they bring. One of our mentors brought our CNC to life and revolutionized our build process, something a student that is only around for 4 years would have trouble achieving. Team 254's mentors bring knowledge to the table as well and I'm very glad they decided to share it. The debate over mentor domination really shouldn't pollute this generous gesture from the poofs.


Also, Thanks for Cheesy Vision! 955 used it with 1 and 2 ball hot at PNWCMP last weekend

Ryan

[notmattlythgoe]

2363 worked on integrating this into our system last night in preparation for the Championship. We were able to get it working with our one ball and will be working on integrating it into our 2 ball tonight.

Thank you for this innovative out of the box system. It is amazing the simple things teams come up with each and every year and how much you can learn by just looking at what other teams have done.

[Coach Norm]

Originally Posted by electrian21

2468 team appreciate used a system like this at Bayou last week. This never occurred to us - it's so simple and elegant. This will be pretty cool to show kids at demos.

Alex, thanks for the shout out on our system including pictures and the code. Kylar has posted an explanation of our system here: http://www.chiefdelphi.com/forums/sh...hreadid=128785

Kylar worked with Greg McKaskle from NI on this implementation. Kylar will be at Championships if you have any questions for him regarding this programming technique.

[rwood359]

Originally Posted by billbo911

Please ask any questions here so I can publicly answer them.

Thanks for posting your receive routine and thanks 254 for the original post. Is there something that we are missing?
We copied your loop into periodic tasks and changed the IP and port. We installed the three routines and changed the IP in CheesyVision. We can't get a connection between the ports. CheesyVision says no connection LabView says error 63 or 65 - connection refused.
Any ideas as to what we are doing wrong?
Thanks

[TikiTech]

A big MAHALO (Hawaiian thank you)!!!

Our team does not have any programming mentors and is entirely student coded. They have been working with the camera tracking the hot goal with some success.

Our students do not use python but were very intrigued by this and had it implemented in a very short time into their C++ code.

In fact our programming cadre has now become INSPIRED to learn more of the language.

THIS IS AWESOME.

Without your gracious sharing of your code I doubt the students would of looked at another programming language, especially this late in the season.

Without a doubt this is what I really love about FIRST. Causing inspiration across the world by sharing.. Keep it up

Good luck to all teams attending the championship!

We will be rocking CheesyVision at St Louis.. See many of you there.

Aloha!

[brunoUC]

Hi, I have been trying to use the Cheesy vision with LabView, but i can't find the correct functions to use it. Is it possible? if so how can i do it?

Thank you!

[billbo911]

Originally Posted by brunoUC

Hi, I have been trying to use the Cheesy vision with LabView, but i can't find the correct functions to use it. Is it possible? if so how can i do it?

Thank you!

As much as I would like to say I can help, I can't.
The solution I posted here will not work! What I posted is a Socket Requester. What is needed is a Socket Receiver.

I have tried dozens of variations based on tutorials, and on-line NI help, but so far have not been able to find anything that will work with CheesyVision and LabView. I know it "should" be simple, but so far I have not found anything that will work.

That said, it could quite easily be my setup. I do not have access to a cRio, so I am using one laptop to run CheesyVision and another running the "receiver" vi in a standalone configuration.

If anyone has any insight in how to resolve this, PLEASE SPEAK UP!!