regarding understanding the clustering result of slic method for a given image

Question

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.

Answer 1

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.