枪械内膛疵病图像的边缘检测算法

An Edge Detector for Inside Flaws in Tubes

首先介绍了零交叉边缘检测算子的基本概念 ,把高斯平滑滤波器和拉普拉斯锐化器结合起来 ,在进行边缘检测的同时平滑掉噪声 ,并提出应用零交叉算子对枪械内膛疵病图像进行边缘检测的实现方法。然后对真实枪械内膛疵病图像进行边缘检测实验 ,结果表明零交叉边缘检测算子检测到的边缘比较完整 ,检测方法直观、简便、运算速度快、边缘定位准确且易于实现。

This paper introduced the basic concepts of a zero-crossing edge-detection operator. Combinimg the Gauss smooth filter with the Laplace lancet window, noises, were smoothed out and the edges were detected. Further, an edge detector for the inside flaws was presented. Comparing the results using zero-crossing edge-detection with results using the Sobel operator edge-detection or the Roberts operator from experiments on the inside flaws in tubes, the zero-crossing edge-detection operator shows more edge-details. The results show that zero-crossing edge detection is simpler, more intuitionistic, faster in calculation, more precise in edge locating and easily implemented.