Observation and Segmentation of Gray Images of Surface Cells in Open Cellular Ceramic Foams.
Jaikoo Park, Sanghun Lee · Journal of the Ceramic Society of Japan · 2001
The surface cells of cellular ceramic foams are analyzed using an optical microscope. The cell density is so abundant that the cell structure can feature a high cell volume fraction, a high degree of interconnection to neighboring cells, variable cell sizes and a high roughness of cell boundaries. To segment these cell or other strut images, a novel image analysis procedure was designed. This includes an application of an algorithm of detecting local maximum regions of light intensity between cells in images and regarding them as strut regions, while the conventional segment operations can only recognize the global maximum or minimum light intensity regions with a threshold intensity value. Then, an effectiveness of the proposed method was compared with earlier typical methods by quantifying the various properties of the cells in each image. As the results, the proposed method shows a good availability in that this gives more efficient presentation of the tiny cells and their strut regions as well as a programmability and reproducibility. However, the improved boundary construction methods are required to enable the proposed method to become more effective.