基于小波和水平集方法的尿沉渣图像分割  

Segmentation of urinary sediment image based on wavelet transform and level set method

作　　者：陈志刚[1] 尹福昌[1] 王斌[1] 

机构地区：[1]长春理工大学光电工程学院,长春130022

出　　处：《计算机应用研究》2008年第9期2878-2880,共3页Application Research of Computers

基　　金：总装备部"十五"预先研究基金资助项目(404050603)

摘　　要：针对尿沉渣图像更为精细的分割,提出了首先使用小波变换和形态学处理消除散焦影响并进行图像的粗分割,然后根据粗分割得到的子图像利用水平集方法找到子图像中目标图像的边缘完成精细分割,为计算机自动识别尿沉渣中的有效成分提供更加准确的依据。实验结果表明,该方法在尿沉渣图像分割的应用中得到了令人满意的结果。A precise segmentation method for urinary sediment image was proposed. Firstly, the wavelet transforms and morphology were used to get rid of the effect of the defocusing and get the subimages that include the particl.es. Then based on the characteristics of the subimages, the level set method was employed felicitously and edge of the particles was located precisely. Finally, the edge data of the particles was processed smoothly applying geometry theory. The experiment results show that the method can segment urinary sediment images effectively and precisely and increasing the performance of urinary sediment image recognition.

关 键 词：小波 水平集 图像分割 尿沉渣 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]