基于粒子滤波和机器视觉的圆形零件尺寸测量  

Measurement for Sizes of Circular Parts Based on Particle Filter and Machine Vision

作　　者：吴新杰[1] 王苗苗[1] 黄国兴[1] 

机构地区：[1]辽宁大学物理学院,沈阳110036

出　　处：《机械设计与制造》2012年第11期49-51,共3页Machinery Design & Manufacture

基　　金：辽宁省自然科学基金项目(20102082)

摘　　要：针对现有基于机器视觉的零件尺寸测量方法在图像受到外界噪声干扰的情况时,测量精度不高的问题,提出一种基于粒子滤波优化算法、机器视觉和MATLAB图像处理技术的圆形零件尺寸测量方法。这里首先简要地介绍基于机器视觉的零件尺寸测量系统的工作原理;然后在分析粒子滤波原理的基础上提出一种基于粒子滤波的优化算法,并介绍利用这种优化算法解决圆形零件尺寸测量的原理和实现步骤;最后进行了仿真实验,仿真实验结果表明:提出的测量方法在零件图像受到外界噪声干扰的情况下,测量精度较高;同时还具有可靠性较高的优点。To deal with low measurement precision of the sizes of parts based on machine vision under the situation of images with noises,a new dimension measurement method of circular parts,which is based on particle filter,machine vision and the technique of MATLAB image processing,is brought forward.Firstly,the basic principle of measurement system of circular parts based on machine vision is briefly introduced.Secondly,after analyzing the basic principle of the particle filtering,a new optimization algorithm based on the particle filter is put forward.Besides,the basic principle and operation procedures which are based on particle filter to measure sizes of circular parts are given.Finally,the simulation experiment is performed.The simulation experimental results have shown that the measurement method here has a higher precision and higher reliability under the situation of images with noises.

关 键 词：粒子滤波 机器视觉 非接触测量 数字图像处理 零件尺寸 

分 类 号：TH16[机械工程—机械制造及自动化] TP391.4[自动化与计算机技术—计算机应用技术]