基于ANFIS的烟丝图像在线检测自适应消噪技术  

On-line Detection of Cut Tobacco Image with Adaptive Noise Cancellation Technology based on ANFIS

作　　者：王晓侃[1] 张艳[1] 

机构地区：[1]河南机电职业学院,河南新郑451191

出　　处：《新技术新工艺》2015年第10期49-52,共4页New Technology & New Process

基　　金：河南省重点科技攻关项目(132102210385;142102210042);河南省教育厅科学技术研究重点项目(13B413968);郑州市科技攻关项目(121PPTGG418)

摘　　要：针对目前烟丝在线检测设备进行图像检测识别时,测量结果容易受到噪声、光线、温度和湿度等的影响,而导致参数时变、时间滞后等问题,设计一种基于MATLAB软件的自适应消噪技术,利用自适应神经模糊推理系统(ANFIS)对烟丝图像噪声的非线性动态特性进行建模,并利用ANFIS逼近有色噪声,然后从测量烟丝图像中消除有色噪声,得到有用的烟丝图像。该技术在实际烟丝在线检测生产过程应用中表明,能够在复杂的环境下,自动地估计出噪声信号,方便、快捷地实现在线去噪,从而得到满意的烟丝图像。Currently when identify the cut tobacco image of on-line detection, the measurement results are easily influ- enced by noise, light, temperature and humidity, and which cause the parameter time-varying and time lag. Design the AN- FIS noise cancellation technology based on the MATLAB software. Modeling the nonlinear dynamic characteristic of noise by ANFIS and approaching the colored noise by ANFIS are presented. Then the colored noise is cancelled from measured signal and the useful cut tobacco image is gotten. This technology is applied in the actual cut tobacco image of on-line detec- tion production process, indicating that it can automatically estimate the noise signal, quickly and easily realize online noise cancellation, thus obtain the satisfied cut tobacco image.

关 键 词：烟丝图像 自适应消噪 模糊神经网络 在线检测 

分 类 号：TN911.7[电子电信—通信与信息系统]