CAJ | 학술논문

本文提出了一种基于模糊数字拓扑的方法来实现医学图象中模糊对象的提取。该方法根据所给定的物体和边缘特征信息将原始医学图象转换到模糊连通空间，获得模糊边界。对复杂形状的对象还提出了模糊边界逐步改进和优化的算法。最终可以跟踪出满足实际需要的连通、平滑及优化的物体轮廓线。算法在超声图象中取得了满意的结果。
본문제출료일충기우모호수자탁복적방법래실현의학도상중모호대상적제취。해방법근거소급정적물체화변연특정신식장원시의학도상전환도모호련통공간，획득모호변계。대복잡형상적대상환제출료모호변계축보개진화우화적산법。최종가이근종출만족실제수요적련통、평활급우화적물체륜곽선。산법재초성도상중취득료만의적결과。
A method for extract fuzzy objects on the basis of fuzzy digital topology. When given characteristic information about the desired object and its edge, the new algorithm transformed the original images into fuzzy connective space and extracted their fuzzy boundary. For objects with complex shape, optimization methods were presented to modify the fuzzy boundary. Experiments on some ultrasonic images showed satisfying results with smoothed, connective and optimal boundaries.