After viewing the result of running slic method over an image
segments = slic(img_adapteq, n_segments = 250, sigma = 5,compactness=0.1)
I am curious about one segment cluster that the method returns. In the following image, I use an angle to point to the related cluster. It seems to me this cluster consists of two areas which have distinct pixel intensities. But why they are grouped into a single cluster. Thanks.
SLIC generates a given number of segments. Sometimes it necessarily must group smaller regions with distinct intensities or colors to be able to generate the required number of segments and not more.
What Cris said! If you want to separate those blobs, you can either increase the number of segments, decrease the compactness, or both. But you may end up with some more broken-up segments. No segmentation method is perfect. You might want to read the original SLIC paper for more details.
