Feature Tracking Tool

Got an idea for something that SharpCap should do? Share it here.
Forum rules
'+1' posts are welcome in this area of the forums to indicate your support for a particular feature suggestion. Suggestions that get the most +1's will be seriously considered for inclusion in future versions of SharpCap.
User avatar
admin
Site Admin
Posts: 13173
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Feature Tracking Tool

#11

Post by admin »

Hi Mike,

glad to hear it worked - I just checked the maths in the code and those are indeed the right values to change.

cheers,

Robin
cairomike
Posts: 46
Joined: Sun Jul 02, 2017 5:18 pm
Contact:

Re: Feature Tracking Tool

#12

Post by cairomike »

Hi Robin

Yes, thanks for the confirmation, it was a fair bet, I compared calibrations West and East and those were the only values that changed sign. I reset the signs again this morning and the cal is still holding true, so I'm very happy to carry on this way until you have the time to implement the mod.

Regards
Mike
cairomike
Posts: 46
Joined: Sun Jul 02, 2017 5:18 pm
Contact:

Re: Feature Tracking Tool

#13

Post by cairomike »

Hi Robin.

I’ve only just caught up with the fact you’d made some changes to the Feature Tracking from version 4.0.9246.

V4.0.9246
Due to sky conditions I haven’t been able to test pier swap, I’ll do that as soon as conditions allow. In the meantime does the feature assume a side of pier when it starts or does it infer it from the saved calibration?

V4.0.9268 & 4.0.9357
You made changes to add minimum contrast control. I can see what you’re doing but it rather negates the major benefit of the feature in as much as it now requires manual intervention to restart guiding after C of M is lost. In previous versions even though it would continue to issue guiding commands and potentially push the mount way off, it did eventually recover and put the Sun back to centre without manual intervention, so it was possible to set image acquisition going and leave it.

In my mind all that’s required from V4.0.9246 is to stop sending guiding corrections but keep monitoring until the C of M was reacquired, i.e while SharpCap is unable to determine where the object is it lets the mount guide uncorrected so it should remain within frame. What do you think?

Regards
Mike
User avatar
admin
Site Admin
Posts: 13173
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Feature Tracking Tool

#14

Post by admin »

Hi Mike,

for the side of pier thing, the code asks the ASCOM driver to report the side of pier when calibrating and then saves that info with the calibration. When you load the saved calibration it also checks the side of pier and flips the movement rate if the two do not match.

For the contrast, I wasn't intending to make you press start again, but what is happening is that the normal code which deals with the tracking lock being lost is kicking in after a set number of frames with no position data. I guess I could either change it to report the same center of mass position repeatedly in case of low contrast or have an adjustment to allow the number of frames before 'tracking lost' to be increased significantly. I think I favour the latter - thoughts?

cheers,

Robin
cairomike
Posts: 46
Joined: Sun Jul 02, 2017 5:18 pm
Contact:

Re: Feature Tracking Tool

#15

Post by cairomike »

Hi Robin.

Thanks for the confirmation of pier side, earlier I looked at the registry and could see the new key.

Between MotoGP races and clouds I’ve managed to verify that pier swap works fine, that's great.

Using version 4.0.9246

It’s been cloudy on and off today so I’ve been able to watch what’s going on.

I can see this is quite tricky, as I’ve been watching the screen today the C of M moves in random ways depending on the direction and thickness of cloud movement. I think your 2nd suggestion is the better of the two, but how about an alternative approach that doesn't need contrast assessment; you have the ‘Dead Zone’ where no corrections are sent, how about an ‘Over of Range’ setting over which you stop issuing corrections in exactly the same way as in the ‘Dead Zone’ whilst still maintaining C of M calculation?

Regards
Mike
User avatar
admin
Site Admin
Posts: 13173
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Feature Tracking Tool

#16

Post by admin »

Hi Mike,

by 'over range', do you mean a 'movement too big' feature, so that if the CoM drifts more than say 50% of the width then the code treats it as false data?

cheers,

Robin
cairomike
Posts: 46
Joined: Sun Jul 02, 2017 5:18 pm
Contact:

Re: Feature Tracking Tool

#17

Post by cairomike »

HI Robin

I was thinking of another area like the 5 pixel 'dead band' say out at 200 pixels which defines the 'over range' limit. So effectively you've got a box where corrections are sent and outside of that nothing is done other than keep monitoring the C of M.

So in pseudo code (teaching my Grandmother tho suck eggs here! :) )

if (Position > DeadBand) and (Position < OverRange) then
ApplyCorrection;
else
DoNothing;

Clearly it would be useful to have the 'over range' limit adjustable, just like the 'dead band'.

From what I'm observing live I think this will work.

Regards
Mike
User avatar
admin
Site Admin
Posts: 13173
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Feature Tracking Tool

#18

Post by admin »

Hi Mike,

Ok, that makes sense - the name needs work though to make it more easily understood :)

cheers,

Robin
cairomike
Posts: 46
Joined: Sun Jul 02, 2017 5:18 pm
Contact:

Re: Feature Tracking Tool

#19

Post by cairomike »

Hi Robin

Just started using V4.09392, which is working fine.

The sky is perfectly clear today so I can't test the latest mods to Feature tracking, ironically. we need a partially cloudy sky which is usually the last thing wanted!!

One observation though; the 'Tracking lost after' setting suggests it's in seconds but the help text says it's frames.

Also, I know my name for the outer zone wasn't quite the right term, but I think the word 'overload' doesn't convey the right meaning either, wouldn't 'Max Move Size'' be better? That also fits better with 'Max Move Duration'.

Last Friday, 30th Sept, I was imaging Jupiter with my C8 c/w AA183C and used the Feature Tracking on C of M, it worked a treat, even though it probably wasn't strictly necessary for the 60s videos I usually capture but a good test, nonetheless.

Regards
Mike
User avatar
admin
Site Admin
Posts: 13173
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Feature Tracking Tool

#20

Post by admin »

Hi Mike,

thanks for the feedback, I will look into both of those issues.

cheers,

Robin
Post Reply