基于MATLAB的CR图像处理在四肢骨边缘检测中的应用
Application of Matlab-based computed radiography images to limb bone edge detection
乔伟 1邱力军2
作者信息
- 1. 第四军医大学生物医学工程系,陕西,西安,710033;解放军第161医院,湖北,武汉,430010
- 2. 第四军医大学生物医学工程系,陕西,西安,710033
- 折叠
摘要
目的:探讨经过Matlab处理后的计算机摄影(CR)图像在四肢骨边缘检测中的应用.方法:利用Matlab工具箱中的Sobel,Roberts,Laplacian,Canny算子对于四肢骨进行边缘检测处理,并比较它们的效果.结果:采用中值滤波处理后,Canny算子获得的图像在骨骼边缘检测的连续性及线条平滑度上效果最好,Sobel算法则在显示骨骼边缘上最直观.直接灰度化处理则均不能满足实际需要.结论:使用Matlab工具箱做四肢骨边缘检测,中值滤波后使用Canny算子得出的边缘检测效果要优于其他处理方法.
Abstract
AIM: To investigate the application of computed radiography (CR) images processed by Matlab to edge-detection of limb bones. METHODS: The edge of limb bone was respectively detected by the Sobel operator, Roberts operator, Laplacian operator and Canny operator from Matlab toolbox and the effects were compared. RESULTS: After median filter processing, the images obtained by Canny operator was the best in bone edge continuity and line smoothness, while the bone edges obtained by Sobel operator were more perceptive. Direct gradation processing could not satisfy the actual need. CONCLUSION: Canny operator from Matlab toolbox is better than other processing methods in limb bone edge detection.
关键词
图像处理/Matlab软件/边缘检测/四肢骨Key words
image processing/Matlab software/edge detection/limb bone引用本文复制引用
出版年
2009
NETL
NSTL
维普
万方数据