摘要
抑制应变片图像采集和传输过程中的噪声,是实现应变片识别和精确安装的关键技术,针对形态学去噪时会导致图像细节模糊的特性,本文引入模糊形态学,根据隶属度来对图像进行划分,将分类的图像像素值分别进行对应的模糊形态学操作,然后在此基础上再对整幅图像进行模糊形态学的开闭平滑来实现对图像的去噪滤波。仿真实验证明该方法可以有效的去除噪声,而且不会对图像的边缘细节造成模糊。
Suppressing the noise of strain gauge in acquisition and transmission is the key in identification and installation. For blurring the details while using the morphology to de-noise. This paper introduces a new algorithm for de-noise the strain gauge image by dividing the image and giving the fuzzy morphology operation, then open and close operation are applied to image to realize de-noise. The simulations in this paper prove that it can perfectly restrain noise and keep image details.
出处
《计算机系统应用》
2010年第12期242-245,共4页
Computer Systems & Applications
基金
国家自然科学基金(60873220)
浙江省自然科学基金(Y108022
Y1090285)
宁波市择优委托计划项目(2009B10003)
关键词
模糊形态学
隶属函数
模糊膨胀
模糊腐蚀
fuzzy morphology
membership function
fuzzy dilation
fuzzy erode
