We propose complex-valued region-based-coupling segmentation neural networks that divide complex-amplitude images into regions and attach labels on them. A neurodynamics is obtained by energy reduction rule in combination with phase-singularity elimination. A millimetric wave radar system is constructed using the network as a preprocessor for image recognition and reconstruction. Experiments demonstrate that the network works as a variable segmentation processor. Relation between network parameters and segmentation functions in the experiments is also reported.