基于图像识别的干涉条纹计数方法研究

Resrarch on Interference Stripe Counting Method Based on Image Recognition

提出了一种基于图像识别方法的干涉条纹计数方法,该方法利用CCD摄像机采集迈克尔孙干涉环图像,采用图像滤波、灰度拉伸、边缘检测、二值化、形态学处理等方法对采集图像预处理,并设计了干涉环识别算法和圆心识别算法两种方法对干涉条纹进行计数,实验结果表明,与传统人工计数方法进行比较,该方法具有测量方便灵活、实时直观,扩展性好等优点,并大大提高了实验数据的测量精度和实验效率,且圆环识别算法比圆心识别算法表现更佳,与波长标准偏差值仅1.4%,满足光干涉测量实际应用.

An interference fringe counting method based on image recognition is proposed.This method uses a CCD camera to collect Michelson interference ring images,and uses image filtering,grayscale stretching,edge detection,binarization,and morphological processing.Image pre-processing was performed,and two methods were designed to count interference fringes:interference ring recognition algorithm and center recognition algorithm.The experimental results show that compared with the traditional manual counting method,this method has the advantages of convenient and flexible measurement,real-time and intuitive,and good scalability And other advantages,and greatly improve the measurement accuracy and experimental efficiency of the experimental data,and the ring recognition algorithm performs better than the circle center recognition algorithm,the standard deviation value is only 1.4%,which meets the practical application of optical interference measurement.