Segment and Feature Extraction of Optical Coherence Tomography Image of Mouse´s Skin In Vivo Using Mathematical Morphology

We applied mathematical morphology to segmentation and characterization of optical coherence tomography (OCT) image of mouse´s skin in vivo. The dilation and the opening operations are used in this study, and the structuring element is important for each operation. We experimentally found that the mouse´s skin was covered with the coverslip for overcoming the irregularity of skin´s surface. And it is demonstrated that preprocessing OCT image of skin with dilation operation, in which the structuring element is line-shaped, could fill the dark area which is due to statistical distribution of photoelectric detecting, and result in segmentation and flattening the skin. Furthermore, we analyzed that the OCT image of mouse´s skin could be composed of various size particle, and opening operation with disk-shaped structuring element was employed to study the particle distribution. The size distribution of OCT image of mouse´s skin could be described as a feature´s characterization.