M.N. Md. Sap, C. Crysdian, and H. Selamat (Malaysia)
Signal and Image Processing, Image Segmentation, Feature Extraction, Medical Image
This paper presents a tree-stage segmentation process for deriving the outer shape of medical image object. The work is motivated by the effort to reduce over segmentation while the image content can effectively be presented. The output would be useful for any content based image analysis system. For that purpose, the improved of the existing techniques have been introduced. Gradient image, watershed transform and region merging are explored and developed for building the complete segmentation algorithm. At last, experiment has been conducted to a collection of medical images, and the result shows that the proposed technique produces the desired output.
Important Links:
Go Back
