News:


  • March 28, 2024, 06:02:39 PM

Login with username, password and session length

Author Topic: VideoF2B at Landres 2018  (Read 4528 times)

Offline Alberto Solera

  • Trade Count: (0)
  • Lieutenant
  • ***
  • Posts: 51
VideoF2B at Landres 2018
« on: August 04, 2018, 09:38:44 AM »
Hello,

Some of you may already know about the program I published few weeks ago.

https://videof2b.blogspot.com/2018/07/presentation.html

I recorded some videos from the fly off at Landres, I think they are really interesting:
(The judges in front of the camera produce some noise, sorry for that)



Please, constructive comments this time...
Best regards  :)

Offline Igor Burger

  • Trade Count: (0)
  • Admiral
  • ******
  • Posts: 2165
Re: VideoF2B at Landres 2018
« Reply #1 on: August 04, 2018, 11:11:04 AM »
Great work with that program Alberto, but why my flight whole from side? :- (((( ... crying :- PPP :- ))) ... you had to tell me, I could move judges little bit :- P

Well, but anyway one comment, I found also myself that taking flights with wide lenses is not usefull, the distorsion is unnatural and watching on screen does not show proper shapes, much better is something about 50mm (for full frame). Unfortunatelly it was not doable this time - whole 2nd round was in wind comming from trees 1m from circle, so longer lenses were not usefull, I hope next year we will see better movies :- )))

Offline john e. holliday

  • 24 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 22752
Re: VideoF2B at Landres 2018
« Reply #2 on: August 04, 2018, 01:34:41 PM »
Wish I hadn't watched this video as they may have been a throw away flight. D>K
John E. "DOC" Holliday
10421 West 56th Terrace
Shawnee, KANSAS  66203
AMA 23530  Have fun as I have and I am still breaking a record.

Online Dave_Trible

  • 24 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 6134
Re: VideoF2B at Landres 2018
« Reply #3 on: August 04, 2018, 05:23:25 PM »
Great videos.  Thank You.  Wish I had such a tool to work with. 

Dave
AMA 20934
FAA Certificate FA3ATY4T94

Offline Randy Cuberly

  • 21 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 3674
Re: VideoF2B at Landres 2018
« Reply #4 on: August 04, 2018, 06:55:49 PM »
Great videos Alberto.  Wonderful program.  Please continue to develop it.

These videos show how much work we as CL stunt fliers actually have to do when in reality the very best in the world still have a long way to go to reach perfection.

The videos also prove beyond any doubt that "Tight" Corners aren't really so tight.  After watching them and taking some measurements on the screen it's obvious that even the "tightest" corners there are about 13 to 14 ft radius and a lot of those flown are closer to an average of 18 ft radius!  Most of the corners are not actually radii at all but irregular curves.

With some different lens technology and computer integration it might be possible to actually develop something like that into a legitimate "Judging program".
Maybe even something like integrating two or three different camera angles into producing a flat image like the pilot is supposed to actually see!  If it's actually flown like that of course!

It certainly is a reasonable future possibility with the right input!

I also think it shows that there just may be more difficult things to fly in a pattern than Corners!

Randy Cuberly
Randy Cuberly
Tucson, AZ

Offline pmackenzie

  • Pat MacKenzie
  • 24 supporter
  • Trade Count: (0)
  • Captain
  • *
  • Posts: 765
Re: VideoF2B at Landres 2018
« Reply #5 on: August 04, 2018, 08:16:21 PM »
Great program!
I didn't have any F2B flights to run through it yet, but just to see what would happen I ran an F3P practice flight video through it.
It had trouble at times due to low contrast between the model and the white gym walls, or when the model slowed down.
But against darker areas it worked quite well.

I have been meaning to anyway, but definitely time to start videoing my F2B flights !

Pat MacKenzie
MAAC 8177

Offline Brett Buck

  • Trade Count: (0)
  • Admiral
  • ******
  • Posts: 13717
Re: VideoF2B at Landres 2018
« Reply #6 on: August 04, 2018, 08:27:35 PM »
Hello,

Some of you may already know about the program I published few weeks ago.

https://videof2b.blogspot.com/2018/07/presentation.html

I recorded some videos from the fly off at Landres, I think they are really interesting:
(The judges in front of the camera produce some noise, sorry for that)


     The system works very well, for the most part, and replicates some of what I see in judging. I think it may not be picking up the finer detail, but excellent for precision assessment.

    I see nothing surprising about the flying, it's the same sorts of issues you always see in any contest. And certainly the corner radius looks like nothing too special, perfectly consistent with the physics. The fact that they don't look like a pure radius is also as expected.

     Brett

Offline Dane Martin

  • 21 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 2804
  • heli pilot BHOR
Re: VideoF2B at Landres 2018
« Reply #7 on: August 04, 2018, 09:29:02 PM »
At my level in flying, I loved watching this. Those triangles were right on top of each other!! I like watching stuff like this, and that tracing helps show how close the expert flyers keep their manuevers!

Offline Air Ministry .

  • Trade Count: (0)
  • Admiral
  • ******
  • Posts: 4978
Re: VideoF2B at Landres 2018
« Reply #8 on: August 04, 2018, 09:35:14 PM »
a half dozen here : https://www.youtube.com/channel/UCnuDKqMogvEwl3yqVbZTsgg

Live real time , would bring the event into the 21st century for the tecno feaks and ' mass media market ' .

Just imagine , thousands of F2B flyers ! .

Could be comparable to the Red Bull Air Races ( Races? even though theyre Aerobatics .  :P )
Though a cool calm commentator would be more bearable than a hysterical twerp , perhaps .

The hysterical twerp is supposed to engender ' Excitement ' .
Mass media market like Yachting , Moto G.P. , etc etc . ! ??

Having looked at a few now , fascinating to see the strengths & weaknesses of the various flyers .

Er JUDGING could be more ' apporpriate ' too . Allowing now generally they get the order pretty much correct without   such innovations .
But to create a 21st century  Mass Market media Event , something of this nature could do that for International Events .
« Last Edit: August 04, 2018, 10:05:59 PM by Matt Spencer »

Offline Keith Renecle

  • Trade Count: (0)
  • Captain
  • *****
  • Posts: 887
Re: VideoF2B at Landres 2018
« Reply #9 on: August 04, 2018, 10:55:08 PM »
Hi Alberto,

Great program! As someone who has traced many flights frame by frame I really appreciate your work and sharing your program. I was just wondering how you save the traced video? I did not see that function in your program, so maybe I missed something? Thanks so much!

Looking at this 2nd round of the fly-off's, even with the side view of Igor's flights, the shapes look similar to what I drew on my CL Sim from 2004, except of the 5 foot corners of course. The added lens distortion is a small problem but even with that, this program is so useful.I am very excited to see this, so many thanks Alberto!

Keith R
« Last Edit: August 07, 2018, 12:07:10 AM by Keith Renecle »
Keith R

Offline pmackenzie

  • Pat MacKenzie
  • 24 supporter
  • Trade Count: (0)
  • Captain
  • *
  • Posts: 765
Re: VideoF2B at Landres 2018
« Reply #10 on: August 05, 2018, 01:47:56 AM »
Hi Alberto,

Great program! As someone who has traced many flights frame by frame I really appreciate your work and sharing your program. I was just wondering you save the traced video? I did not see that function in your program, so maybe I missed something? Thanks so much!

Hi Keith,

    it saves a copy as an avi in the same folder as the original video.  Same file name as the original but with "_out.avi" at the end.

Pat
MAAC 8177

Offline katana

  • Trade Count: (0)
  • Commander
  • ****
  • Posts: 161
Re: VideoF2B at Landres 2018
« Reply #11 on: August 05, 2018, 05:50:43 AM »
As not a great fan of aerobatics, that program / tracing made the videos far more watchable, giving an accurate descriptive representation of what is flown. One thing I would change is the trace decay time as it sometime exceeded the two laps between and old and new patterns got muddled - other than that excellent work.

Offline Keith Renecle

  • Trade Count: (0)
  • Captain
  • *****
  • Posts: 887
Re: VideoF2B at Landres 2018
« Reply #12 on: August 05, 2018, 05:55:53 AM »
Hi Keith,

    it saves a copy as an avi in the same folder as the original video.  Same file name as the original but with "_out.avi" at the end.

Pat
Aah.....thanks Pat! I'll have a look at that.
Keith R

Offline EddyR

  • Trade Count: (0)
  • Admiral
  • ******
  • Posts: 2562
Re: VideoF2B at Landres 2018
« Reply #13 on: August 05, 2018, 06:48:46 AM »
  I have been a couch since the days Bill Rich and I flew together and one thing jumped out at me was only one flyer stood so he was facing the plane. Most were flying the plane to there right.
  I remember comments about the worlds many years ago that the “corners were soft”.  It looks like they listened.
  I wrote a article on here a few years ago about what the eye and the brain is expecting can be a lot different than what is happening. There are some good articles on Google about it.
   Bill and I filmed many flights and soon learned that body position was one of the most important parts of practice.
 
EddyR
Locust NC 40 miles from the Huntersville field

Offline Chris_Rud

  • 2014 Supporters
  • Trade Count: (0)
  • Commander
  • *
  • Posts: 189
Re: VideoF2B at Landres 2018
« Reply #14 on: August 05, 2018, 07:39:24 AM »
How does one get this program? Looks like a good coaching tool.

-ChrisRud

Offline Igor Burger

  • Trade Count: (0)
  • Admiral
  • ******
  • Posts: 2165
Re: VideoF2B at Landres 2018
« Reply #15 on: August 05, 2018, 07:43:08 AM »
How does one get this program? Looks like a good coaching tool.

-ChrisRud

The link is in first mail  ;D

Offline Brett Buck

  • Trade Count: (0)
  • Admiral
  • ******
  • Posts: 13717
Re: VideoF2B at Landres 2018
« Reply #16 on: August 05, 2018, 10:48:52 AM »
How does one get this program? Looks like a good coaching tool.

   I agree. I still think you want a real coach, too, because this is not picking up on small details, but for shape and intersections, it looks like just what we try to draw by hand.

    Brett

Offline Alberto Solera

  • Trade Count: (0)
  • Lieutenant
  • ***
  • Posts: 51
Re: VideoF2B at Landres 2018
« Reply #17 on: August 05, 2018, 12:06:55 PM »
Thank you all for your comments!

First the doubts:
The download link is in my blog, in the first message. There are some problems in windows with some antivirus detecting it as a virus (hard to solve).
The video with the trace is saved in the same folder, as mentioned.

If it is the general wish I will make the line shorter. You can delete it pressing space, but I forgot to do it in some videos. Maybe I could make a "maneuver detection" to delete the previous one.

Igor, sorry for your video, you had too many spectators at the right spot!  ;) It works better placing the camera at 30-35 meters from the center, but with so many people around it was worse this time.
Anyway, this camera is already calibrated, so the only effect is perspective, straight lines are preserved in the image. (I went to calibrate it only to find it was already calibrated  ::))

This morning I have tested a real time solution. Using a raspberry pi board, I can process the video and send it to several smartphones or laptops thought wifi with negligible delay. It still needs some work but opens new possibilities.

I think I would be able to draw a 3D trajectory using more cameras, but I think the setup would get too complicated, we will see...

Best regards  :)

Offline Brett Buck

  • Trade Count: (0)
  • Admiral
  • ******
  • Posts: 13717
Re: VideoF2B at Landres 2018
« Reply #18 on: August 05, 2018, 12:47:49 PM »
I think I would be able to draw a 3D trajectory using more cameras, but I think the setup would get too complicated, we will see...

   Just from geometry, you would only need two cameras, and carefully survey their orientation and relative position. Then check it with reflective markers or something. You also need some sort of a synchronous timing arrangement - it doesn't have to be accurate in absolute terms (I don't recommend GPS) but you need something to keep them the same time base within milliseconds.

   If you know any model rocket/"spacemodeling" guys, there have been extensive analyses of a two-station altitude tracking setup that is exactly what you need. Each discrete sample of your cameras gives you and azimuth and elevation vector from the camera to the airplane, from that and knowledge of the camera position and orientation, you can get a position of the airplane in your system's "reference frame" in 3D. That is precisely what model rocket tracking station data reduction does, the difference being they do it once to get a single 3D position fix and then, finally, an absolute altitude reference. You care about is the entire fix, but you also only care about relative to the other fixes you get once per 5 milliseconds, or something.  That sounds daunting but with the proper implementation, child's play for it to be done in the 10-ish seconds you have before the next maneuver.

    It would be a simple matter to add lens distortion to the system, correct each vector based on where the vector falls in the field of view. Most photographic manipulation software already does this, with data from the lens manufacturer, and it is also sensed by the camera and done on the fly when creating "processed" formats inside the camera.

    Once you have that, you can easily generate the track from other viewpoints for coaching purposes.

   I also think you will need some sort of targets on the airplane to ensure you never lose track (as it does sometimes in the videos). If it means better coaching or or more accurate scores, you won't have any trouble convincing people to do that, I expect.

   Still and all, very good and offers great potential.

     Brett




Online Crist Rigotti

  • 23 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 3855
  • Electric - The future of Old Time Stunt
Re: VideoF2B at Landres 2018
« Reply #19 on: August 05, 2018, 04:20:33 PM »
All I can say is amazing!
Crist
AMA 482497
Waxahachie, TX
Electric - The Future of Old Time Stunt

Online Dave_Trible

  • 24 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 6134
Re: VideoF2B at Landres 2018
« Reply #20 on: August 05, 2018, 07:34:34 PM »
I downloaded the link and hope to try recording with the Sony camcorder, then download to the laptop.  Had to order some Sony-specific cables for HDMI.  If that doesn't work I will see if I can find a tripod for the IPhone.....I'm excited to try this!

Dave
AMA 20934
FAA Certificate FA3ATY4T94

Offline Alberto Solera

  • Trade Count: (0)
  • Lieutenant
  • ***
  • Posts: 51
Re: VideoF2B at Landres 2018
« Reply #21 on: August 06, 2018, 06:07:07 AM »
   Just from geometry, you would only need two cameras, and carefully survey their orientation and relative position. Then check it with reflective markers or something. You also need some sort of a synchronous timing arrangement - it doesn't have to be accurate in absolute terms (I don't recommend GPS) but you need something to keep them the same time base within milliseconds.

   If you know any model rocket/"spacemodeling" guys, there have been extensive analyses of a two-station altitude tracking setup that is exactly what you need. Each discrete sample of your cameras gives you and azimuth and elevation vector from the camera to the airplane, from that and knowledge of the camera position and orientation, you can get a position of the airplane in your system's "reference frame" in 3D. That is precisely what model rocket tracking station data reduction does, the difference being they do it once to get a single 3D position fix and then, finally, an absolute altitude reference. You care about is the entire fix, but you also only care about relative to the other fixes you get once per 5 milliseconds, or something.  That sounds daunting but with the proper implementation, child's play for it to be done in the 10-ish seconds you have before the next maneuver.

    It would be a simple matter to add lens distortion to the system, correct each vector based on where the vector falls in the field of view. Most photographic manipulation software already does this, with data from the lens manufacturer, and it is also sensed by the camera and done on the fly when creating "processed" formats inside the camera.

    Once you have that, you can easily generate the track from other viewpoints for coaching purposes.

   I also think you will need some sort of targets on the airplane to ensure you never lose track (as it does sometimes in the videos). If it means better coaching or or more accurate scores, you won't have any trouble convincing people to do that, I expect.

   Still and all, very good and offers great potential.

     Brett

Theory is clear, but practice...
I collaborated in a research project where a couple of stereo cameras were used to detect UAVs position (it didn't work). The two cameras were specific for machine vision and connected via Ethernet for synchronization. I remember they were expensive.

The system would only be usable in large competitions, with a specialized operator and some investment... I would be happy to spend a year working on the project with adequate funding, but being realistic I don't find enough justification at the moment. I also like to build and fly from time to time  :)
Thank you for the input anyway!

Next update will improve the computation time a lot and include some minor fixes. Next could be the Raspberry implementation and then maybe a mobile app.

 H^^

Offline Keith Renecle

  • Trade Count: (0)
  • Captain
  • *****
  • Posts: 887
Re: VideoF2B at Landres 2018
« Reply #22 on: August 06, 2018, 07:40:43 AM »
Alberto, your program is great even as it is. Thanks so much! The next step would be a system that plots the 3-D co-ordinates without camera's but of course this is also complex and takes much time for somebody to do this without funding. We are  looking at this possibility here in South Africa and will do some basic tests soon. This of course means adding a sensor system to the model but for a serious judge training program I think that this is justified. This could be used for judging as well but that software would be a big challenge and I also realize that it could take many years for a system to get accepted by the majority. This however is the future to really judge who the true champion is. For now, your program is a giant leap forward for training!

Keith R
Keith R

Offline Igor Burger

  • Trade Count: (0)
  • Admiral
  • ******
  • Posts: 2165
Re: VideoF2B at Landres 2018
« Reply #23 on: August 06, 2018, 09:10:46 AM »
For now, your program is a giant leap forward for training!

May be if Alberto can post those videos without tracking then it could be good to take the maneuver without thrackinh to allow "judging" and then to show tracking to allow the judge to test himself if he saw what is displayed by computer.

Offline Keith Renecle

  • Trade Count: (0)
  • Captain
  • *****
  • Posts: 887
Re: VideoF2B at Landres 2018
« Reply #24 on: August 06, 2018, 10:34:34 AM »
That's a good idea!  y1
Keith R

Offline Igor Burger

  • Trade Count: (0)
  • Admiral
  • ******
  • Posts: 2165
Re: VideoF2B at Landres 2018
« Reply #25 on: August 06, 2018, 10:44:10 AM »
And actually I got dirty idea, we can make program testing judges consistency by showing them many maneuvers, let them judge and reapeat some of them, then we can see tendency to fading, balooning etc ... well looks like I will have troubles to ge to the finals after this message :-P  VD~

Offline Keith Renecle

  • Trade Count: (0)
  • Captain
  • *****
  • Posts: 887
Re: VideoF2B at Landres 2018
« Reply #26 on: August 06, 2018, 10:50:20 AM »
Well......that's the end of Igor  n1
Keith R

Offline Randy Cuberly

  • 21 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 3674
Re: VideoF2B at Landres 2018
« Reply #27 on: August 06, 2018, 11:44:49 AM »
Well......that's the end of Igor  n1

I know you guys are joking of course, and there may actually be a few who might take that attitude.  However I think it's a great idea and I believe most of the guys I know that judge would be very receptive to the Idea.  A good judge should not mind being put on the spot to learn and to explain his judgement.  NO certainly not during a competition but most assuredly during training sessions.  How else can one really learn.  Here in Tucson we often have sessions similar to that at the judges warm up flights.  After the flight everyone compares their scores and those that are very different than the NORM are asked to explain their reasons for choosing those numbers.

Typically, the guys that fly the warm up flight are asked to deliberately make several errors during the flight (yes that's probably not necessary, but we do it anyway).
The guys that do these flights are typically pretty good expert level fliers who are either not entered in that class or not entered in the contest.  We have no problem coming up with duffers here in Tucson that fit that description.

It works, and I for one believe it really helps establish a commonality between the judges.  I haven't experienced any complaints over doing it that way!

After all Judging is supposed to be done to a standard of rules, not opinions!  Ideally I think (although I have no idea how to do it) the identity of the flier would be kept from the judges until after the contest!  (again that's probably impossible but would insure no Halo effect)

Again I certainly know you guys were kidding I just wanted to make this statement!

Randy Cuberly

Randy Cuberly
Tucson, AZ

Offline Alberto Solera

  • Trade Count: (0)
  • Lieutenant
  • ***
  • Posts: 51
Re: VideoF2B at Landres 2018
« Reply #28 on: August 06, 2018, 12:39:48 PM »
May be if Alberto can post those videos without tracking then it could be good to take the maneuver without thrackinh to allow "judging" and then to show tracking to allow the judge to test himself if he saw what is displayed by computer.

Good idea

https://www.youtube.com/playlist?list=PLMDN8gK5NWCorkiVRKK6MULFP-rBVR1zU

The other one is all yours  :D

Keep us informed Keith, if I can help don't doubt to contact me.

It is certainly a great tool for judges. I trained for a week with my father as coach and we both made a nice step forward. He felts much more confident about his observations after that.
It is also better for us, now I can see how bad I have flown without him telling me!  ::)

Alberto

Offline Air Ministry .

  • Trade Count: (0)
  • Admiral
  • ******
  • Posts: 4978
Re: VideoF2B at Landres 2018
« Reply #29 on: August 06, 2018, 09:08:10 PM »
Judging aside , . . .

Makes the event decipherable to anyone .

TELEMETRY would achive similar result .

Obviously may need some development as to concentricity ! . ( big word ) .
but could allow a ' true track ' of hypothetical ( rule book ) manouvre, also.
as a Datum , which may mean evolution of flightpath to true track . If it isnt now . ? .

Obviously a wondrous contrivance , Alberto . Yr to be congradulated etc .

Offline Igor Burger

  • Trade Count: (0)
  • Admiral
  • ******
  • Posts: 2165
Re: VideoF2B at Landres 2018
« Reply #30 on: August 06, 2018, 10:38:15 PM »

Offline Keith Renecle

  • Trade Count: (0)
  • Captain
  • *****
  • Posts: 887
Re: VideoF2B at Landres 2018
« Reply #31 on: August 07, 2018, 12:00:28 AM »
I wasn't joking Randy! If Igor was out in Landres I would have moved further up the ranks.......instead of 61st place........I would have been 60th!  #^

I still believe that there was some kind of conspiracy by the French organizers. My team mate Evert Scholtz and I had number 60 and 61 as competitor numbers, and that's how we placed. Sounds fishy to me!  <=

Keith R
Keith R

Offline Alberto Solera

  • Trade Count: (0)
  • Lieutenant
  • ***
  • Posts: 51
Re: VideoF2B at Landres 2018
« Reply #32 on: August 15, 2018, 10:53:32 AM »
There is a new version with some fixes in the blog:

http://videof2b.blogspot.com/2018/08/new-release-v03-and-updates.html

Best regards  :)

Offline Randy Cuberly

  • 21 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 3674
Re: VideoF2B at Landres 2018
« Reply #33 on: August 15, 2018, 02:46:32 PM »
I wasn't joking Randy! If Igor was out in Landres I would have moved further up the ranks.......instead of 61st place........I would have been 60th!  #^

I still believe that there was some kind of conspiracy by the French organizers. My team mate Evert Scholtz and I had number 60 and 61 as competitor numbers, and that's how we placed. Sounds fishy to me!  <=

Keith R


See, I told you!  You can't trust the French!   LL~ LL~ LL~ LL~ LL~ LL~ LL~ LL~ LL~ LL~ LL~ LL~ LL~ LL~

The Swiss...Maybe!  but not the French!   n~ n~   LL~ LL~ LL~ LL~

Randy Cuberly
Randy Cuberly
Tucson, AZ


Advertise Here
Tags:
 


Advertise Here