机器视觉疲劳裂纹扩展试验摄像头自动聚焦方法

The auto-focusing method for the camera in machine vision fatigue crack propagation test

摘要:

在机器视觉疲劳裂纹扩展试验中,为了能够满足对裂纹宏观观察裂纹和准确定位裂纹尖端的要求,需要采用变焦镜头放大和缩小采集图像范围,针对于摄像头变焦后的聚焦问题,提出了摄像头的自动聚焦方法。首先建立图像采集系统,采集裂纹图像至计算机,通过中值滤波对图像预处理,去除噪声干扰,选取裂纹区域作为聚焦窗口,采用Laplace算子法作为清晰度评价函数,并提出一种变步长穷举法进行聚焦搜索。最后设计了以ARM7为核心的摄像头运动控制器。实验表明,所提出的方法能够实现疲劳裂纹扩展试验中摄像头在各种情况下的准确自动聚焦,为下一步精确测量疲劳裂纹扩展参数奠定基础。

Abstract:

In the machine vision fatigue crack propagation test, in order to observe the crack macroscopically and locate the crack tip accurately, the zoom lens was needed to enlarge and narrow the image. For the problem that the camera needed to be re-focusing after zooming, an auto-focusing method was proposed. After establishing the image acquisition system, the crack image was collected to computer, and then median filter was used to pre-processing the image to remove the noise. The crack area was selected as the focusing window and evaluated sharpness of the area by Laplace operator method, and then an adaptive step size exhaustive method was used to search the focus. The controller based on ARM7 was designed to control the motion of camera. The research results show that for all situations the proposed method can realize the accurate auto-focusing of the camera in the fatigue crack propagation test, and it built the foundation to accurately measure the fatigue crack growth parameters.