摘要
为了克服返波管太赫兹连续波成像系统所获取图像存在的对比度低、干涉条纹强等缺陷,较好地区分出目标和背景,避开不规则干涉条纹的干扰,采用基于模糊局部信息C均值的聚类算法用于目标检测,并针对太赫兹图像性质,对其中的隶属度函数进行了改进。结果表明,新的聚类分割算法适用于具有不规则条纹干扰下的太赫兹连续波图像,能较好地提取出图像中的目标,较经典图像聚类算法具有更好的目标检测精度。
In order to overcome the backwards,such as low contrast and strong interference fringes of the images captured by terahertz( THz) continuous wave imaging system based on backward wave oscillator,to detect target objects from background and avoid the disturbance of the irregular interference fringes,the cluster method based on C-means of fuzzy local information was used to target detection and the membership function was improved so that it was suitable to terahertz images. Experiment results show that the proposed clustering algorithm can detect target objects from terahertz images corrupted with irregular interference fringes and has better accuracy than the classic image clustering algorithm.
出处
《激光技术》
CAS
CSCD
北大核心
2015年第3期289-294,共6页
Laser Technology
基金
国家自然科学基金资助项目(61371160
60971059)
关键词
图像处理
太赫兹连续波成像
聚类
模糊局部信息
隶属度加权
image processing
THz continuous wave imaging
clustering
fuzzy local information
membership weight