基于机器视觉的尼龙齿轮检测研究

针对综合验光仪中的尼龙齿轮的工业检测精度和效率要求,应用一种新的齿轮实际分度圆直径获取方法,设计基于视觉技术的尼龙齿轮检测系统。利用CCD相机采集齿轮图像,通过Halcon软件对图像进行预处理,以计算分度圆直径作为基准圆,然后往大、往小每隔0.1个像素生成一个新的圆,共计51个圆,由此计算出每个圆上的齿厚和齿槽宽,获得每组数据的方差,将所有圆的方差建立数组,方差最小的圆即为实际分度圆。从而获得实际分度圆直径。齿轮实际分度圆直径获取方法具备较好的准确率和稳定性。设计的基于视觉技术的齿轮检测系统能够满足工业的生产需求。

Study on Nylon Gear Detection Based on Machine Vision

In order to meet the requirement of industrial detection accuracy and efficiency of nylon gear in synthetic optometry,a new method for obtaining the actual dividing circle diameter of gear was proposed,and a nylon gear detection system based on visual technology was designed.By using CCD camera gear image,through the Halcon software of image preprocessing,circles as a benchmark to calculate dividing circle diameter,then small to large,to every 0.1 pixels to generate a new round,a total of 51 round,thus calculated each circle on the tooth groove width,thickness and the variance of each group of data obtained,will build an array of all round variance,variance of the smallest circle is the actual dividing circle.Thus,the actual diameter of the dividing circle can be obtained.The result is that the method of obtaining the diameter of the actual dividing circle of the gear has better accuracy and stability.The conclusion is that the gear detection system based on visual technology can meet the industrial production needs.