摘要
为了辅助医生仅能依赖肉眼观察CT图像从而判断CT影像好坏,同时改善在医学图像的分割算法上的性能,提高分割的效率,本文提出了一种使用遗传算法和KSW熵法相结合的CT图像肺部分割方法,该方法通过对医学图像的预处理、构建基于遗传算法的KSW熵方法寻求最优的图像分割解以实现对肺部实质区域的提取、使用改进的Ostu算法获得较好的自适应阈值以实现图像的二值化、通过图像形态学进行图像修补、使用区域生长法对肺部实质性区域的左右部分进行分离。实验结果表明,本文提出的算法在对医学肺部图像的分割时在分割精度上比其他医学图像分割方法都有了一定程度的改善,具有较强的分割自适应性和稳定性。
Regularly,doctors observed CT images by naked eyes to judge whether they were good or not.In order to assist them,meanwhile,to improve segmentation algorithms performance in medical images,this paper proposed the combination of genetic algorithm and KSW entropy method in CT lung image segmentation.The medical image was processed from pre-processing,KSW based on genetic algorithm method was used to get the best segmentation solution for the extraction of lung parenchyma region,the improved Ostu was used in image binaryzation,and image morphing was used for image inpainting,regional growth was used to extract the left and right parts of the pulmonary parenchyma region.The testing results illustrate that the algorithm proposed in this paper has a certain improvement in the segmentation precision than the other methods,and has a strong segmentation self adaptability and stability.
作者
姚立平
潘中良
YAO Liping;PAN Zhongliang(School of Physics and Telecommunications Engineering,South China Normal University,Guangzhou 510006,China)
出处
《电视技术》
2018年第11期1-6,共6页
Video Engineering
基金
广州市科技计划项目
广东省科技计划项目(2016B090918071
2014A040401076)
作者简介
姚立平(1994—),硕士研究生,研究方向:基于机器视觉的图像处理;潘中良(1966-),博士,教授,研究方向:计算机应用、嵌入式系统设计。
