Poster presentation: The segmentation method for the boundary detection of lesion area using a levelset method based on real-coded genetic algorithm (MEとバイオサイバネティックス)

概要

論文の詳細を見る
When the doctor does diagnosis, the quality of ultrasound images is poor to recognize the lesion area because of effects on poor contrast, blurred edges, etc. The purpose is to reduce noises and edge detection in ultrasound image. In this paper, the border curve in edge segmentation is represented by using the level set method to detect the tumor region of human liver on the ultrasound image. We preprocessed the image by using a real-coded genetic algorithm (RCGA) based on speckle reducing anisotropic diffusion (SRAD). As the result, we confirm to get the dominant segmented image which has more distinctive and smooth character than images which are used previously.
2008-01-18