基于粒子群算法的工件视觉测量研究

The research on mechanical measure and recognition system based on machine vision

提出了一种基于机器视觉的测量和识别系统。在自动化生产中,测量技术在产品检查、装配和品质管理占有重要的地位。其中视觉系统可以说是众多测量工具最具潜力一种。与其他系统相比,视觉系统的最主要优点是具有高速、高精度且能够方便的调整测量范围。首先对采集的图像进行预处理;其次利用模板匹配及图像分割技术获得所需要测量的二值图像;最后利用粒子群算法对其进行模式聚类并计算出特征具体的尺寸参数。结果表明该系统能够正常工作、具有很高的工程应用价值。

It proposed a measurement and recognition system based on machine vision. In automated manufacturing, measurement is important in inspection,assembly and quality assurance. One of the most desirable measurement tools is a vision system. The main advantage of employing vision systems over other approaches are easily modified measuring margin,high speed and high accuracy .First,pre-processing algarithms was used to enhance the quality of origin image. Second, The template matching algorithm and Image segmentation algorithm for the image processing are utilized to obtain the binary image ;Final,the PSO algorithm was utilized to pattern cluster and calculate the specific size parameters. The result revealed that measurement vision systems could be realized using industrial camera and has high value in engineering applications.