Mean shift based gradient vector flow for image segmentation.

Highlights: [•] Classical GVF based algorithms face challenges such as over- or under-segmentation. [•] Our MSGVF algorithm pursues a minimised combination of mean shift and GVF terms. [•] The proposed algorithm is evaluated against two publicly accessible databases. [•] MSGVF achieves better segmentation performance than the classical methods. [ABSTRACT FROM AUTHOR]