Figure 3

Qualitative segmentation results of four methods for kidney in US image. (a) original images; (b) manual segmentations (in green); (c) DRLSE method segmentations; (d) DRLSE segmentations on the NLTV denoised image; (e) the result of our method without using the alignment model after the segmentation; (f) the result of our method with using the alignment model after the segmentation.