基于模糊区域生长的医学图像分割
Medical Image Segmentation Based on Fuzzy Region Growth
郭晓蓉 1任小玲1
作者信息
- 1. 西安工程大学计算机科学学院 西安 710048
- 折叠
摘要
针对肺部图像的特征,论文提出了一种基于模糊隶属度的区域生长的图像分割方法.首先,对原图像进行预处理,并将其二值化;其次根据二值图像确定不同区域轮廓,计算不同区域的质心;再次,根据标注图像自动获取目标区域的质心将其作为区域生长的种子点;最后利用隶属度函数计算种子点邻近像素点的隶属度,根据其隶属度的大小进行区域生长.实验结果表明,论文提出的方法与传统区域生长方法相比能够缓解过分割或者欠分割现象,分割精度比传统方法提高了21.88%,可以分割出更清晰的目标区域,为今后的临床诊断提供帮助.
Abstract
According to the characteristics of lung image,a region growing image segmentation method based on fuzzy mem-bership degree is proposed in this paper.First of all,the original image is preprocessed and binarized.Secondly,the contours of dif-ferent regions are determined according to the binary image,and the centroids of different regions are calculated.Thirdly,according to the labeled image,the centroid of the target region is automatically obtained as the seed point of the region growth.Finally,the membership function is used to calculate the membership degree of the seed point adjacent to the pixel point,and the region growth is carried out according to its membership degree.The experimental results show that,compared with the traditional region growing method,the proposed method can alleviate the phenomenon of over-segmentation or under-segmentation,the segmentation accura-cy is 21.88%higher than the traditional method,and the target area can be segmented more clearly.It provides help for future clini-cal diagnosis.
关键词
图像分割/区域生长/模糊隶属度/最大隶属度/质心Key words
image segmentation/region growth/fuzzy membership degree/maximum membership degree/centroid引用本文复制引用
出版年
2024
国家科技期刊平台
NSTL
万方数据