{ "id": 6800, "title": "Random black line on an image?", "url": "/competitions/galaxy-zoo-the-galaxy-challenge/discussion/6800", "author_name": "", "post_date": "2014-01-10T12:04:15.830Z", "votes": null, "comment_count": 10, "views": 2117, "content": "

This is 108474.jpg in test images. I saw a line through the galaxy. Could there be more of these in image sets?

", "messages": [ { "id": "37217", "postDate": "01/10/2014 12:04:15", "content": "

This is 108474.jpg in test images. I saw a line through the galaxy. Could there be more of these in image sets?

", "rawMarkdown": "", "votes": null }, { "id": "37218", "postDate": "01/10/2014 12:05:28", "content": "

This relates to Question 3 in the decision tree: "Q3. Is there a bar?".

", "rawMarkdown": "", "votes": null }, { "id": "37219", "postDate": "01/10/2014 12:10:55", "content": "

Thanks for the answer. I thought this was some kind of a mistake by image producer. So I suppose the bar does not have to go through (or near) the center of the galaxy?

", "rawMarkdown": "", "votes": null }, { "id": "37220", "postDate": "01/10/2014 12:12:46", "content": "

Actually, the question is "Is there a sign of a bar feature through the centre of the galaxy?". You can see all the questions here.

", "rawMarkdown": "", "votes": null }, { "id": "37221", "postDate": "01/10/2014 12:14:02", "content": "

This isnt a bar, this is an artifact. It is the separation between CCDs on the camera that took the photo and the galaxy happens to fall on it. 

\n

 

\n

My guess there would be more of these like this and it is unavoidable and unfortunately a fact of the data

", "rawMarkdown": "", "votes": null }, { "id": "37222", "postDate": "01/10/2014 12:14:58", "content": "

I think by 'bar' they mean galactic bars, like this for example. I've noticed that image distortions like this one are not uncommon in the dataset (a lot of tears as well), I think this may be related to question 8, answer 3 ('disturbed' galaxies) but I'm not sure, and I haven't tried to verify this so far.

\n

 

", "rawMarkdown": "", "votes": null }, { "id": "37223", "postDate": "01/10/2014 12:17:37", "content": "

[quote=AstroDave;37221]

\n

This isnt a bar, this is an artifact. It is the separation between CCDs on the camera that took the photo and the galaxy happens to fall on it. 

\n

My guess there would be more of these like this and it is unavoidable and unfortunately a fact of the data

\n

[/quote]

\n

Wow okay, that changes some of the decision I will be taking then! I thought I was supposed to catch these bars in question 3. As sedielem writes, this might be captured in question 8 though.

", "rawMarkdown": "", "votes": null }, { "id": "37224", "postDate": "01/10/2014 12:19:50", "content": "

sedielem is correct, the bar we are looking for is the central yellow thing in the picture he attached

", "rawMarkdown": "", "votes": null }, { "id": "37225", "postDate": "01/10/2014 13:33:06", "content": "

[quote=sedielem;37222]

\n

I think by 'bar' they mean galactic bars, like this for example. I've noticed that image distortions like this one are not uncommon in the dataset (a lot of tears as well), I think this may be related to question 8, answer 3 ('disturbed' galaxies) but I'm not sure, and I haven't tried to verify this so far.

\n

 [/quote]

\n

There's publicly available material - a presentation by Kyle Willett, at an astronomy conference, if I recall correctly* - which strongly suggests that that's exactly what a lot of zooites did (classify a galaxy with the 'black line' artifact running through it as 'disturbed' in question 8).

\n

 

\n

Needless to say, that's not what was intended! A galaxy may appear 'disturbed' because, for example, it's highly asymmetric but still has a general 'spirally' shape (many post-merger galaxies look like this); and 'overlaps' (chance projections, where two galaxies far apart in physical space appear to overlap) often look like disturbed galaxies (there is no 'overlaps' button/option in question 8; an oversight perhaps).

\n

 

\n

* if anyone's interested, I'll see if I can find a link to this, or similar

", "rawMarkdown": "", "votes": null }, { "id": "37226", "postDate": "01/10/2014 13:39:24", "content": "

The reason I was considering this is because the schematic image for a 'disturbed' galaxy in the decision tree at the bottom of this page looks like it has a tear in the middle, and such image tears (as well as other artifacts, like black lines) are not uncommon in the training data. But based on what you're saying I guess question 8 does not have anything to do with imaging artifacts. Good to know :)

\n

 

\n

EDIT: after a closer look at the (new) training data, I can confirm that the galaxies considered most 'disturbed' by the zooites almost invariably have one of these black lines running across them.

", "rawMarkdown": "", "votes": null }, { "id": "37235", "postDate": "01/10/2014 16:40:31", "content": "

[quote=sedielem;37226]

\n

EDIT: after a closer look at the (new) training data, I can confirm that the galaxies considered most 'disturbed' by the zooites almost invariably have one of these black lines running across them.

\n

[/quote]

\n

So it's not a bug, it's a 'feature', right? ;-)

", "rawMarkdown": "", "votes": null } ], "comments": [ { "id": 37218, "author_name": "utdiscant", "author_url": "", "post_date": "01/10/2014 12:05:28", "content": "

This relates to Question 3 in the decision tree: "Q3. Is there a bar?".

", "votes": null, "replies": [] }, { "id": 37219, "author_name": "jwh143479", "author_url": "", "post_date": "01/10/2014 12:10:55", "content": "

Thanks for the answer. I thought this was some kind of a mistake by image producer. So I suppose the bar does not have to go through (or near) the center of the galaxy?

", "votes": null, "replies": [] }, { "id": 37220, "author_name": "utdiscant", "author_url": "", "post_date": "01/10/2014 12:12:46", "content": "

Actually, the question is "Is there a sign of a bar feature through the centre of the galaxy?". You can see all the questions here.

", "votes": null, "replies": [] }, { "id": 37221, "author_name": "astrodave", "author_url": "", "post_date": "01/10/2014 12:14:02", "content": "

This isnt a bar, this is an artifact. It is the separation between CCDs on the camera that took the photo and the galaxy happens to fall on it. 

\n

 

\n

My guess there would be more of these like this and it is unavoidable and unfortunately a fact of the data

", "votes": null, "replies": [] }, { "id": 37222, "author_name": "sedielem", "author_url": "", "post_date": "01/10/2014 12:14:58", "content": "

I think by 'bar' they mean galactic bars, like this for example. I've noticed that image distortions like this one are not uncommon in the dataset (a lot of tears as well), I think this may be related to question 8, answer 3 ('disturbed' galaxies) but I'm not sure, and I haven't tried to verify this so far.

\n

 

", "votes": null, "replies": [] }, { "id": 37223, "author_name": "utdiscant", "author_url": "", "post_date": "01/10/2014 12:17:37", "content": "

[quote=AstroDave;37221]

\n

This isnt a bar, this is an artifact. It is the separation between CCDs on the camera that took the photo and the galaxy happens to fall on it. 

\n

My guess there would be more of these like this and it is unavoidable and unfortunately a fact of the data

\n

[/quote]

\n

Wow okay, that changes some of the decision I will be taking then! I thought I was supposed to catch these bars in question 3. As sedielem writes, this might be captured in question 8 though.

", "votes": null, "replies": [] }, { "id": 37224, "author_name": "astrodave", "author_url": "", "post_date": "01/10/2014 12:19:50", "content": "

sedielem is correct, the bar we are looking for is the central yellow thing in the picture he attached

", "votes": null, "replies": [] }, { "id": 37225, "author_name": "jeantate", "author_url": "", "post_date": "01/10/2014 13:33:06", "content": "

[quote=sedielem;37222]

\n

I think by 'bar' they mean galactic bars, like this for example. I've noticed that image distortions like this one are not uncommon in the dataset (a lot of tears as well), I think this may be related to question 8, answer 3 ('disturbed' galaxies) but I'm not sure, and I haven't tried to verify this so far.

\n

 [/quote]

\n

There's publicly available material - a presentation by Kyle Willett, at an astronomy conference, if I recall correctly* - which strongly suggests that that's exactly what a lot of zooites did (classify a galaxy with the 'black line' artifact running through it as 'disturbed' in question 8).

\n

 

\n

Needless to say, that's not what was intended! A galaxy may appear 'disturbed' because, for example, it's highly asymmetric but still has a general 'spirally' shape (many post-merger galaxies look like this); and 'overlaps' (chance projections, where two galaxies far apart in physical space appear to overlap) often look like disturbed galaxies (there is no 'overlaps' button/option in question 8; an oversight perhaps).

\n

 

\n

* if anyone's interested, I'll see if I can find a link to this, or similar

", "votes": null, "replies": [] }, { "id": 37226, "author_name": "sedielem", "author_url": "", "post_date": "01/10/2014 13:39:24", "content": "

The reason I was considering this is because the schematic image for a 'disturbed' galaxy in the decision tree at the bottom of this page looks like it has a tear in the middle, and such image tears (as well as other artifacts, like black lines) are not uncommon in the training data. But based on what you're saying I guess question 8 does not have anything to do with imaging artifacts. Good to know :)

\n

 

\n

EDIT: after a closer look at the (new) training data, I can confirm that the galaxies considered most 'disturbed' by the zooites almost invariably have one of these black lines running across them.

", "votes": null, "replies": [] }, { "id": 37235, "author_name": "jeantate", "author_url": "", "post_date": "01/10/2014 16:40:31", "content": "

[quote=sedielem;37226]

\n

EDIT: after a closer look at the (new) training data, I can confirm that the galaxies considered most 'disturbed' by the zooites almost invariably have one of these black lines running across them.

\n

[/quote]

\n

So it's not a bug, it's a 'feature', right? ;-)

", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "37217": "", "37218": "", "37219": "", "37220": "", "37221": "", "37222": "", "37223": "", "37224": "", "37225": "", "37226": "", "37235": "" }, "source": "meta" }