Automated Video Tracking of Planes

What could be of interest is south Wales, UK?;)

Actually, this was discussed privately a year ago, but, regrettably, never made a public thread.
I can understand scientists wanting more evidence, but you would think the Chemians would jump
What could be of interest is south Wales, UK?;)

Actually, this was discussed privately a year ago, but, regrettably, never made a public thread.

For (any) fans of [SHCC], [RCT] is coming, but the spec. is proving ridiculousDSCF2809.JPG .. Here is the birth of the main drive:
 
Last edited by a moderator:
That looks very cool, except for the name, which would break the politeness policy here.

SHCC is also very cool.
I did wonder about that. I could change it by deed poll.
I aim to track to the diffraction limit of the telesope over a second or so, and an absolute accuracy of about 100th of a degree.
I have a simple mount for the camera which tracks smoothly, but the mems yaw giro difts badly, so it can't be left do do its own thing.

ps.
The idea behind the [tangent] geometry is that the ball joint between the ballscrew and telescope pivot and linear bearings follows a scaled down version of path of the aircraft. Luckily for us that means nearly constant speed, heading and altitude. An alt/az drive would mean an infinite slew rate for an aircraft overhead where it is most photogenic.
 
Last edited:
I did wonder about that. I could change it by deed poll.
I aim to track to the diffraction limit of the telesope over a second or so, and an absolute accuracy of about 100th of a degree.
I just Google'd SHCC and saw some YouTube videos - they're awesome. Is it overlaying the flight info in real time, if so did you create the software and are you planning on releasing it?

Ray Von
 
I just Google'd SHCC and saw some YouTube videos - they're awesome. Is it overlaying the flight info in real time, if so did you create the software and are you planning on releasing it?

Ray Von
Yes, I get the adsb data direct from the aircraft as there would be unknown delays if I hacked FR , for instance. I use dump1090 and Rasberry pi. Older versions using the PC and other software used up too much processor. I used a cheap dvbt? dongle for most of those but have gone over to the very good Flightaware dongle, filter, and antenna. I get so much data my software now crashes after a while..

I don't plan to release it _ publicly_ as the Labview licensing situation is unclear, to say the least.

The overlayed graph needs to be aligned, scaled, rotated etc. to fit known landmarks, but I could in principle set it up and run here for any IP webcam covered by my adsb receiver, or any other that can be accessed. I could then think about compiling a version for somebody.

I am using a foscam pzt webcam, but could use any I could find a command for to return a jpg. It worked ok with my phone as a webcam.

The videos are screen captures.

It is quite a lot of fiddling to do, but I am open to the general idea of sharing.
 
Last edited:
Son of RCT.jpg
Awesome! I want one!
This one is just a gimbal mount, of the type used for drones, I made for my camera. SHCC enables me to pick a target and sends it to another VI (Labview program) which sends estimated elevation and azumuth data via bluetooth to the mount at a rate of100Hz.
http://www.olliw.eu/storm32bgc-v1-wiki/Main_Page
I plan to add an optical shaft encoder to cure the drifting yaw azis.
 
Last edited:
I've been developing similar software, more for automated video tracking of rocket launches.


It should be possible to do similar kinds of plane watching with it, but I haven't tested it for that lately. Integrating ADS-B data for plane watching would be cool, but to be honest I don't know much about the hardware side of things or how to capture that data. I'm using an off-the-shelf telescope with an Alt-Az mount which doesn't even give direct drive rate control in the command set (don't ask me how I pulled it off). Obviously that does have the limitation of not being able to follow close to the zenith. It's a natural fit for tracking satellites though since the telescope already knows (and can be polled for) its celestial coordinates.
 
My understanding is that optic tracker does support ADS-B based acquisition for tracking, but it is not compatible with my telescope (Meade LX200 Classic). In fact as far as Meade goes it only supports the newer autostar LX200s, nothing else. That's what originally motivated me to start writing it. I'm coding mine to support a much wider range of Meade telescopes (and hopefully include support for Celestrons as well).
 
I've been developing similar software, more for automated video tracking of rocket launches.


It should be possible to do similar kinds of plane watching with it, but I haven't tested it for that lately. Integrating ADS-B data for plane watching would be cool, but to be honest I don't know much about the hardware side of things or how to capture that data. I'm using an off-the-shelf telescope with an Alt-Az mount which doesn't even give direct drive rate control in the command set (don't ask me how I pulled it off). Obviously that type sites does have the limitation of not being able to follow close to the zenith. It's a natural fit for tracking satellites though since the telescope already knows (and can be polled for) its celestial coordinates.


Great stuff. It would be good if someone did it with javascript, HTML, or something. I too bought an Alt-Az mount, but so far, controlling via a PC is useless as the documentation is poor and it wants to do silly things like a crash stop before you can change speed. Can you imagine a car like that! There is plenty of information on getting the ADSB data with a ~£5 dongle and free software on most of the FlightRadar-type sites and others like http://www.satsignal.eu/raspberry-pi/dump1090.html

There is an HTML steam which is easy to use.
A 'Raw' hex stream which is a bit of a pain but contains most information.
A 'Basetation steam' port 30003 which I use mainly because the Dump1090? puts a high resolution time stamp on it.
 
Last edited:
Another approach is the Optic Tracker software controlling a "GoTo" telescope

Source: https://www.youtube.com/watch?v=6v3J6MSkTgs


I wonder if combining the two approaches might be useful, although real-time ADS-B based tracking seems pretty solid as-is.


I got an Optic Tracker trial but did not do much with it. My telescope cam is a very nice chinese webcam that will stream HD over IP. I think the Optic Tracker insisted on USB. THis was done with the RCT telescope and said cam from one mile:
Source: https://www.youtube.com/watch?v=1qMvl_BA7ao

Look, no hedge!

Probably this cam: http://www.elpcctv.com/hd-1080p-mini-atm-ip-camera-with-36mm-hd-lens-p-62.html
 
Last edited:
My understanding is that optic tracker does support ADS-B based acquisition for tracking, but it is not compatible with my telescope (Meade LX200 Classic). In fact as far as Meade goes it only supports the newer autostar LX200s, nothing else. That's what originally motivated me to start writing it. I'm coding mine to support a much wider range of Meade telescopes (and hopefully include support for Celestrons as well).

I think there is some open-source universal language for telescopes (used by optic Tracker?) but I need to look into that again. My mount is an Skywatcher Allview, but basically I understand that a lot of these cheaper telescopes are made by the same chinese OEM. for Celestrons etc. so I could maybe test your software for you :)
 
I think there is some open-source universal language for telescopes (used by optic Tracker?) but I need to look into that again. My mount is an Skywatcher Allview, but basically I understand that a lot of these cheaper telescopes are made by the same chinese OEM. for Celestrons etc. so I could maybe test your software for you :)
I believe you're referring to ASCOM. I don't think Optic Tracker uses ASCOM, and neither could I. I tried it at first, but quickly ran into problems. ASCOM wasn't designed to support high speed tracking like this. It might be possible to use the direct drive rate commands with ASCOM for the telescope mounts that will support that command, but mine will not. The method I'm using right now simply doesn't work through ASCOM due to throughput limitations; ASCOM forces the telescope to acknowledge the completion of the last command before it will accept a new command.
 
I believe you're referring to ASCOM. I don't think Optic Tracker uses ASCOM, and neither could I. I tried it at first, but quickly ran into problems. ASCOM wasn't designed to support high speed tracking like this. It might be possible to use the direct drive rate commands with ASCOM for the telescope mounts that will support that command, but mine will not. The method I'm using right now simply doesn't work through ASCOM due to throughput limitations; ASCOM forces the telescope to acknowledge the completion of the last command before it will accept a new command.

That's the one! Thanks for the info.
 
FlightAware have tarted up the local web server for Dump 1090, although some things don't seem to work, and they broke SHCC by changing the format. I think it is more appealing to most peopleflightaware dump 1090.PNG .
 
Just a note to say that I have not given up. It is going to work, but once I know that I tend to lose interest. I spent many weeks messing with oils and damping, which was never in the plan. The main surprise there was that greases have stiction, and silicone greases are oils with 'fillers' like fumed silica, which are extremely abrasive, like microscopic eggshells. I ended up making my own by filling viscous silicone oil with graphite. Small amounts of compliance, or friction can bypass damping once the amplitude drops to low levels. The main killer is that it can never be completely autonomous, because some aircraft do not transmit accurate enough data. There seem to be about half a dozen different avionics systems in use, with different accuracies, and approaches to dither. Possibly late adopters have better systems. Delta are one of the best. I attach a very few examples plots from different aircraft systems, showing speeds derived from individual position transmissions .aal 705.PNG baw183.PNG clx767.PNG cfg2ec.PNGdlh462.PNG ryr663l.PNG
 
Last edited:
FR509 /RYR41LK Ryanair Type(B738) Boeing 737-8AS Registration EI-DCX
I think.
Agree. I've just checked the FR24 playback for the given time and date :cool: The B738 did a short flight from Bristol to Dublin and needed not to climb higher. I vaguely remember an earlier Metabunk post on a similar observation.

Added:
Contrails from jet liner at lower than usual altitudes have been discussed here:
https://www.metabunk.org/persistent-cumuliform-contrails-at-25-000-feet-over-california.t6299/
 
Last edited:
I watched BAW50 climbing from about 28,000 to 36,000 without any visible trail at all. I am guessing it was an aerodynamic trail, but the lighting at the time, and also the auto-colour balance, DWDR and auto-exposure tends to make everything look grey unless you have some nice white clouds in the frame.

I did not look outside to see if it persisted. The webcam with a 20mm. dia. achromat is looking through a window upstairs, and controlled remotely via Ethernet. The standard goto mount is not good enough to use the telescope, or track near vertical, mainly because it takes a second or three to send and receive position data. I am making a complete replacement.
 
Here's the latest results of my tracking software. I used manual joystick control due to the very dynamic nature of SpaceX's stage separation and boost back burn. The viewfinder video would be too confusing to the tracking algorithm, and if I'm going to have to manually control the tracking I prefer to get a feel for it at the start rather than at a critical moment like staging.

Nevertheless, the LX200 did a fantastic job smoothly tracking it with my joystick commands. I went off the main camera view on an external monitor.
 
I'm just using the standard LX200 Classic mount, no physical modifications.

Just found out about your Pre-Alpha release of TeleTrak Windows PC software for the Meade LX200 a couple days ago (4/22/18) and was able to successfully download from your Astronomy Live YouTube channel provided URL (
Source: https://youtu.be/eDUfJxsuUPk
).

Been looking for such a program, while other app is available, the price range is some what costly.

So far, I installed and ran it on my Win7 x64 PC with a USB cam for within the app visual tracking testing purposes. Got an error of "The program can't start because opencv_gpu2410.dll is missing from your computer. Try reinstalling the program to fix the problem." but TeleTrak was able to run.

Downloading OpenCV-2.4.10.exe (https://sourceforge.net/projects/opencvlibrary/files/opencv-win/2.4.10/opencv-2.4.10.exe/download) to install to see if the error will go away.

Your use of color and contrast based tracking detection is simple and yet very effective. Was able to track a green (color based detection) eye drop bottle quickly. After the object was completely out of the camera view and when brought back in to the camera view, the tracking resumed by isolating the green color. WoW!! Now that's pretty neat. ;0)

I have several GoTo telescopes (Meade LX200GPS 12", Celestron Nextstar 8", Meade ETX 60AT/70AT/80AT) and would be glad to be part of your testing and evaluation team.

Let me know how I can assist and ways to provide feedback and suggestions if such evaluation team exist. Where do I sign up?

Next step is to connect the Meade LX200GPS AutoStar II to the PC via RS232/USB and mount the USB cam securely to the telescope to see if the scope will move using TeleTrak. Once successful, the next fun step is to track moving objects like planes, trains and automobiles. More planes than anything else. :0)

Keep up the AWESOME work and looking forward to future release updates of TeleTrak.
 
Last edited:
So far, I installed and ran it on my Win7 x64 PC with a USB cam for within the app visual tracking testing purposes. Got an error of "The program can't start because opencv_gpu2410.dll is missing from your computer. Try reinstalling the program to fix the problem." but TeleTrak was able to run.

Downloading OpenCV-2.4.10.exe (https://sourceforge.net/projects/opencvlibrary/files/opencv-win/2.4.10/opencv-2.4.10.exe/download) to install to see if the error will go away.

Your use of color and contrast based tracking detection is simple and yet very effective. Was able to track a green (color based detection) eye drop bottle quickly. After the object was completely out of the camera view and when brought back in to the camera view, the tracking resumed by isolating the green color. WoW!! Now that's pretty neat. ;0)

I have several GoTo telescopes (Meade LX200GPS 12", Celestron Nextstar 8", Meade ETX 60AT/70AT/80AT) and would be glad to be part of your testing and evaluation team.

Let me know how I can assist and ways to provide feedback and suggestions if such evaluation team exist. Where do I sign up?

Next step is to connect the Meade LX200GPS AutoStar II to the PC via RS232/USB and mount the USB cam securely to the telescope to see if the scope will move using TeleTrak. Once successful, the next fun step is to track moving objects like planes, trains and automobiles. More planes than anything else. :0)

Keep up the AWESOME work and looking forward to future release updates of TeleTrak.

Thank you so much! Thanks for the heads up about the missing library error, let me know if it causes any issues with the program aside from the error itself. I'm planning to port the program to python if I can find a suitable GUI framework, preferably something as easy to set up as visual studio and is compatible with python 3. Seems like most options are restricted to python 2.7. Also, I haven't had an opportunity to properly test it on an Autostar telescope, but I'm pretty certain it will not work on Celestron at the moment. Let me know how it goes with your Autostar, but be prepared to shut it down if it starts behaving erratically.

I'm keen to make an option in the Python version of the program to use ASCOM for mounts that support arbitrary values for the MoveAxis method in ASCOM. That excludes most Meade telescopes, even my own LX200 classic, but newer autostar LX200's support it I think, as do most Celestron mounts and other high end mounts. I do have concerns about lag induced by passing the commands through ASCOM; in my initial experiments with it, it seemed pretty slow compared to commanding the telescope through direct serial commands. Still, it would at least let me instantly support a much wider array of telescopes, even if the performance was a bit slower.
 
View attachment 32656
Thank you so much! Thanks for the heads up about the missing library error, let me know if it causes any issues with the program aside from the error itself. I'm planning to port the program to python if I can find a suitable GUI framework, preferably something as easy to set up as visual studio and is compatible with python 3. Seems like most options are restricted to python 2.7. Also, I haven't had an opportunity to properly test it on an Autostar telescope, but I'm pretty certain it will not work on Celestron at the moment. Let me know how it goes with your Autostar, but be prepared to shut it down if it starts behaving erratically.

I'm keen to make an option in the Python version of the program to use ASCOM for mounts that support arbitrary values for the MoveAxis method in ASCOM. That excludes most Meade telescopes, even my own LX200 classic, but newer autostar LX200's support it I think, as do most Celestron mounts and other high end mounts. I do have concerns about lag induced by passing the commands through ASCOM; in my initial experiments with it, it seemed pretty slow compared to commanding the telescope through direct serial commands. Still, it would at least let me instantly support a much wider array of telescopes, even if the performance was a bit slower.

This may be old news but just incase.

----------

https://docs.python.org/3/faq/gui.html

What platform-independent GUI toolkits exist for Python?

Depending on what platform(s) you are aiming at, there are several. Some of them haven’t been ported to Python 3 yet. At least Tkinter and Qt are known to be Python 3-compatible.

Tkinter
Standard builds of Python include an object-oriented interface to the Tcl/Tk widget set, called tkinter. This is probably the easiest to install (since it comes included with most binary distributions of Python) and use. For more info about Tk, including pointers to the source, see the Tcl/Tk home page. Tcl/Tk is fully portable to the Mac OS X, Windows, and Unix platforms.

Qt
There are bindings available for the Qt toolkit (using either PyQt or PySide) and for KDE (PyKDE4). PyQt is currently more mature than PySide, but you must buy a PyQt license from Riverbank Computing if you want to write proprietary applications. PySide is free for all applications.

Gtk+
The GObject introspection bindings for Python allow you to write GTK+ 3 applications. There is also a Python GTK+ 3 Tutorial.

----------
Content from External Source
I'm currently learning Python v3 right now and may just look up the different GUI
platform availability. I may attempt to replicate your existing Teletrak GUI in Python
once I get to the appropriate level of mastering Python. :0). Got to set a high milestone once in a while in one's life.
[off topic chat removed]
 
Last edited by a moderator:
Back
Top