Joint deconvolution/segmentation of microscope images of materials

In this paper, we propose the joint deconvolution and segmentation of materials images by incorporating blurring information in the EM/MPM segmentation algorithm. In the segmentation of microscope images of materials, exact boundary precision is very important. But it is difficult to get good results if the images have some degradation obtained in the acquisition process. We incorporate prior knowledge of blurring degradation into the existing EM/MPM segmentation algorithm in order to improve segmentation results at object boundaries. Experimental results using materials datasets are presented to demonstrate the proposed method is effective for that purpose.