K-Means clustering based bi-level coarse image segmentation

The proposed technique K-Means clustering based bi-level coarse image segmentation performs location-specific segmentation on any given input image and produces the binary image of the object of interest. This technique allows us to interactively define the boundaries of the region of interest (ROI) and produces the coarse image of that ROI. As the segmentation process is confined to the subregions of the given image, this technique promises accuracy and reduced computational time.