How to separate the foreground from the background in an image?

Question

Rachel 2014-1-4

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/111403-how-to-separate-the-foreground-from-the-background-in-an-image

编辑： Image Analyst 2014-1-4

Hi,

I am working on a project related to image matting. I am trying to separate the foreground of an image from the background. Can anybody please help?

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Image Analyst 2014-1-4

Most likely, if you supply more information such as an image and a description of what is the foreground and what is the background and what you want to measure (area, count, perimeter, color, etc.)

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Image Analyst 2014-1-4

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/111403-how-to-separate-the-foreground-from-the-background-in-an-image#answer_120027

See my Image Segmentation Tutorial, BlobsDemo: http://www.mathworks.com/matlabcentral/fileexchange/?term=authorid%3A31862

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

Answer 2

Walter Roberson 2014-1-4

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/111403-how-to-separate-the-foreground-from-the-background-in-an-image#answer_120028

There are a number of "optical illusions" around that prove that there is no objective meaning to "foreground" and "background". What is "foreground" in one interpretation might be "background" in a different interpretation of the same image.

3 个评论
显示 1更早的评论隐藏 1更早的评论

Walter Roberson 2014-1-4

I am thinking of a couple of the usual suspects, Necker Cubes, "The Young Lady and the Old Crone", the Chalice.

Looking around briefly, the hollow Chaplin mask is good

I had not seen the Necker Lattice before.

A few others at http://en.wikipedia.org/wiki/Ambiguous_image

Image Analyst 2014-1-4

编辑：Image Analyst 2014-1-4

one eyed 3D - 1.jpg

Thanks for those. By the way, there is some research into "Main Subject Detection" - I think by those involved in stockphotography databases (CBIR) and camera manufacturers:

http://www.bing.com/search?q=main+subject+detection&qs=n&form=QBLH&pq=main+subject+detection&sc=0-0&sp=-1&sk=&cvid=2c077cda730d4ae3a447c74ab97b7343

https://www.google.com/#q=main+subject+detection

Also, attached is an example that I've collected of one-eyed 3D. Blow it up full screen, it looks fairly planar with both eyes open, then close one eye and it looks more 3D-like than if both eyes are open.

请先登录，再进行评论。