I have a raster (.jpg) which I have segmented utilising Segment Mean Shift to smooth out JPEG artefact (aim is to classify it as accurately as possible for conversion into vectors).
Spectral Detail: 13
Spatial Detail : 15
Minimum Segment Size In Pixels: 20
Are there specific limitations or input parameters that I have missed ?
In addition, when applying a Maximum Likelihood Classification I get this result: