Timeline for Remove background from image
Current License: CC BY-SA 3.0
12 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| May 23, 2017 at 12:41 | history | edited | CommunityBot |
replaced http://stackoverflow.com/ with https://stackoverflow.com/
|
|
| Nov 18, 2013 at 8:07 | history | bounty ended | CommunityBot | ||
| Nov 15, 2013 at 18:03 | comment | added | dmsnell | @rockinfresh - thanks for the algorithm, but for real images this will rarely work. The images must be perfectly aligned in order to be useful and they must have the same size. In most photographs, we usually have different orientations. Alignment algorithms can take care of this for us by matching up images with different lighting, perspective, and size. Secondly, because your algorithm simply checks if the background is darker than the foreground, it makes unwarranted assumptions about the background. What happens if the foreground is dark and the background bright? | |
| Nov 14, 2013 at 5:45 | vote | accept | Darshana | ||
| Nov 14, 2013 at 5:45 | comment | added | Darshana | this method is very slow. then I found this blog: Fast Pixel Operations in .NET (With and Without unsafe). | |
| Nov 12, 2013 at 18:51 | comment | added | Abid Rahman K | @rockinfresh: What I wanted to say is, while answering, first explain the algorithm, then provide its code (as given in the second answer here). Most of the time, readers won't be that experienced or even newbies in the field of image processing. So the explanation will help them. | |
| Nov 12, 2013 at 18:45 | comment | added | rockinfresh | @AbidRahmanK, My algorithm flow and idea is pretty simple and straightforward. Take background image(1) and background image with foreground objects(2). Compare pixel by pixel. If the pixel are the same, make it to black, eliminating it. Whatever remains is the odd ones out, basically the foreground objects. Hope that helps (: It's a very common method. | |
| Nov 12, 2013 at 18:40 | comment | added | rockinfresh | the .3, .59 and .11 are the RGB luminance value. You can find out more from these links: scantips.com/lumin.html and groups.google.com/forum/#!topic/xsi_list/bv3y0WX2R9Q. While this is supposedly the default, there are many people who uses their own values too. Hope that helps(: | |
| Nov 12, 2013 at 18:30 | comment | added | Abid Rahman K | please explain your algorithm first. | |
| Nov 12, 2013 at 6:12 | comment | added | Darshana |
what are these .3, .59 and .11?
|
|
| Nov 11, 2013 at 19:32 | review | First posts | |||
| Nov 11, 2013 at 19:55 | |||||
| Nov 11, 2013 at 19:17 | history | answered | rockinfresh | CC BY-SA 3.0 |