How to find the mean color value of superpixels (over-segmentation) for 3D label image ?

If I have a labeled image resulting from an over-segmentation method similar as SLIC superpixels method for 2D image. However, I have a 3D image. Now my question is how to replace the label value by the mean color value in all 3D dimension? I found in MATLAB documentation how to apply it for 2D (https://www.mathworks.com/help/images/ref/superpixels.html) but when I apply each view independently the results are not consistent and may change across views. for that reason how can I replace the value of superpixels by the mean value of the image?

Answer 1

[yanqi liu]

yes,sir，may be use for loop to split 3D image as 2D image list，and use mean2 on every image，finally rerange to 3D result output

[MatlabUser]

I think this also will generate inconsistent results, I already figured it out thanks!