Adjacent frames
-
by Ptd
Is it possible to get at the eight adjacent patches of sky so an asteroid can be tracked as it passes into the frame next door?
Posted
-
by interplanetary scientist
This is a possible future effort we're considering. Was there a case of that in this frame?
Posted
-
by Ptd
I'm not sure it was this one I was referring to, it was the image I was looking at directly before I started the discussion, I think I did that rather than post a comment in the object itself, if so my mistake sorry, and, unless I've made a comment on an image so it shows up in my profile, I can't see a way of looking back through images I've previously classified, to id the culprit for you. Is there a way?
Sorry Ptd.
Posted
-
by itsVance
I am new here. I found a frame with multiple objects moving in different direction across all 4 frames. I want to mark them all, but it seems to only want me to mark 1 object for each set of 4 frames. How do I do that? Sorry to hijack your post here. I just cant seem to figure out how to start one of my own =)
Posted
-
by Ptd
I had that problem Vance, but I think I found the answer, finish tracking one particular object through all four images, clicking not visible if necessary on some of those frames; then click "done" the interface then lets you mark another series.
Posted
-
by Ptd
Found another one where, if it is real then there will be appearances in adjacent frames.
Image AAZ00009b7
If one of the science team glances at this post I have a technical question. When the data set is chopped up for classification on this site, is there an overlap built in, so that if there is something interesting at the very edge of the screen, that isn't cropped in all the images that object will appear in? Reason for asking is, if the answer is yes, then I can probably bin this thread.
Many thanks
PtdPosted
-
Well I'm not a member of the science team, but I have the power of experience.
On many occasions I've observed asteroids to slip off of the image during the course of the frames, so as far as I can tell there's nothing preventing from such a thing for happening. In fact, if they could identify moving objects with such certainty, there wouldn't be much point for an asteroid-hunting crowdsourcing project, would there?
Posted
-
by Dr.Asteroid scientist, admin
There is overlap built into the frames - if something slips off, we are guaranteed enough overlap that it would appear on another set. Even a very fast moving NEO would still be captured.
Posted
-
So when it says 'When playing through frames look for movement among the aligned images, preferably in at least 3 of the images.' in the guide, that's not including the special case when an object is only in 2 of the 4 frames?
Posted