中文摘要：

本文提出一种以LabVIEW为软件平台结合IMAQ Vision函数库的钢球表面缺陷检测方法,该方法完成了钢球表面缺陷的有无与类型判断。首先,利用Daubechies4小波的分解、最佳树计算、阈值量化与重构,对钢球图像进行去噪处理。其次,利用钢球图像纹理的熵进行有无缺陷的判断。在此,本文提出了基于水平熵、垂直熵、对角线熵的综合熵作为评判参数。然后以形状因子F对有缺陷的钢球图像分类。最后,在搭建的硬件实验台上测试了检测软件,满足快速、准确的现场要求。

英文摘要：

Taking LabVIEW as the software platform together with IMAQ vision function library,this paper put forward a detection method of steel ball surface defect.With this method,we could determine whether there were surface defects or not,and identify the types of surface defects.Firstly,it was to reduce noise of the steel ball image through decomposition of Daubechies4 wavelet,calculation of optical tree,quantization of threshold and reconfiguration.Secondly,we determined whether there were surface defects or not...