MAR Vision Day [Archive]

laura1640

06-10-2016, 17:26

Hello MAR Teams,

I am pleased to announce an upcoming event open for all MAR teams called MAR Vision Day! This will be a collaborative workshop held on November 12th in Downingtown, Pennsylvania for teams to improve their vision processing code.

For those of you who may have never heard of vision processing, it is the idea of using a camera to control the robot. We will be talking about using vision processing to automatically align the robot with the high goal.

This workshop is designed for teams with all levels of experience. Whether you have never done any vision processing, or you already have a working vision system, this is a workshop for you! We hope to have all teams collaborate as we all work to improve our vision code. There will be a hands-on part of the workshop, so teams are welcome to bring their robot. If your team is interested in attending, please fill out the following form, and we will contact you with more information.

https://goo.gl/forms/ZN2tCkoieoEXdfjS2

TL;DR Team 1640 (Oh My Lordy!) will be hosting a MAR Vision Day on November 12th

EmileH

06-10-2016, 20:30

Will the presentations be live streamed so us New England folk can benefit from this seminar?

Overall it sounds like a great idea!

NShep98

06-10-2016, 20:46

Will the presentations be live streamed so us New England folk can benefit from this seminar?

Overall it sounds like a great idea!

+1

Seeing as I will turn 18 on that day, it seems as good a gift as any. Plus, I get to nag my team about camera use again.

Brian Maher

06-10-2016, 21:15

I will pass this along to 1257, though you may have a hard time getting New Jersey teams with Brunswick Eruption being on the same day.

laura1640

08-10-2016, 16:58

Will the presentations be live streamed so us New England folk can benefit from this seminar?

Overall it sounds like a great idea!

We had not considered live streaming the event. I am not sure how well that would work since we were planning to have this event as more of a hands-on workshop than a presentation.

However, I like the idea. We will certainly give it some thought. There may be some parts we can live stream.

+1

Seeing as I will turn 18 on that day, it seems as good a gift as any. Plus, I get to nag my team about camera use again.

Happy almost birthday @NShep98

Turing'sEgo

08-10-2016, 18:55

We had not considered live streaming the event. I am not sure how well that would work since we were planning to have this event as more of a hands-on workshop than a presentation.

If you don't have the means to livestream it, you could always record it then post it later.

Eric222

05-11-2016, 13:07

What will be the hours of the workshop?

dradel

05-11-2016, 22:36

would be great if it could be recorded and posted! even more so if you are going to cover using lab view as well. my programers have been working on vision and made some progress, but more info is always great.

Greg Hainsworth

06-11-2016, 06:43

Have any details been blasted out yet about this workshop? We haven't heard anything yet.

Thanks.

laura1640

06-11-2016, 20:54

We just sent an email to all registered teams. If you did not receive one, please let us know.

would be great if it could be recorded and posted! even more so if you are going to cover using lab view as well. my programers have been working on vision and made some progress, but more info is always great.

If by LabView, you mean NI Vision, then yes. There will be teams coming with NI Vision experience that I am sure would love to answer any questions you may have.

kmckay

14-11-2016, 13:21

Thanks again for having us from team 5401

laura1640

28-11-2016, 21:11

Here is the link to the presentation. Let me know if you have any questions :)

https://www.youtube.com/watch?v=2EboyI9zAUM

NShep98

28-11-2016, 21:43

Is there a link to the actual slideshow? It's a bit difficult to see.

SamCarlberg

29-11-2016, 00:32

About the raw OpenCV stuff vs. GRIP (starts around 17:00): the next release has code generation for the OpenCV 3.1.0 APIs for Java, C++, and Python 3. You can set up the pipeline and tune parameters in GRIP, then generate a class that does all the OpenCV functions that you can use in your robot program on the roborio, or wrap in a program running on a coprocessor. You don't need to write any vision processing code, but you will need to write the code for getting images to the generated code and using the OpenCV objects it outputs.

laura1640

29-11-2016, 23:27

Is there a link to the actual slideshow? It's a bit difficult to see.

Absolutely. Here is the link:

https://docs.google.com/presentation/d/1VvIJnuHq1RzO3emtlbjhFHbet55Smm8_4AqWav65jpI/edit?usp=sharing

About the raw OpenCV stuff vs. GRIP (starts around 17:00): the next release has code generation for the OpenCV 3.1.0 APIs for Java, C++, and Python 3. You can set up the pipeline and tune parameters in GRIP, then generate a class that does all the OpenCV functions that you can use in your robot program on the roborio, or wrap in a program running on a coprocessor. You don't need to write any vision processing code, but you will need to write the code for getting images to the generated code and using the OpenCV objects it outputs.

This sounds like a very interesting feature. Thank you.