How did you get the .pbm pictures?

[baizhenmao95]

How did you get the .pbm pictures?

[debidatta]

Hi,

There are ways to do foreground segmentation (finding pixels of the salient object in the image) using computer vision. If you are willing to click on object of interest you can use GrabCut (https://github.com/opencv/opencv/blob/master/samples/python/grabcut.py). If you want a more automated solution, you can take a look at this. This is slightly dated. So, may be look at something a little more new: this. I'll recommend using Bilinear pooling for cleaning masks. Don't worry about the .pbm extension. Once you get a binary mask as an image you can just store it as a pbm using OpenCV. Hope this helps.

[baizhenmao95]

Thank you very much! I'll have a try.