基于小波变换的手骨图像边缘提取方法

Method of Picking up the Image Edge of Hand Bone on the Basis of Wavelet Transformation

为了探求可以同步分离手指骨和腕骨各部分轮廓的方法,便于骨龄测定。针对医学图像特点和要求,引入小波变换,结合传统图像分割方法,采用边缘生长和寻找大区域的方法,实现手骨同步分离。研究表明该方法能够较为精确的、全自动的分离手骨各部分,能够同步分离目标边缘,同时保留细节信息。

To explore a method which can segment the carpal bones and phalanges simultaneously based on the wavelet technique, and combining with the methods of edge growing and looking for big area, the method can realize accurate and full-automatic segmentation. This algorithm can avoid setting threshold. And better result is achieved than the traditional method. This scheme can promote the automation of computer-assisted bone age assessment system.