CAJ | 학술논문

首先应用数学形态学方法进行图像滤波,然后对Sobel边缘检测算子进行改进,并引入最小二乘拟合方法进行算法求精,从而构造出一种新的理论体系完备的子像素级边缘检测算法,从滤波、边缘定位、拟合求精等多个方面对传统的边缘检测算法加以改进,使其可以在诸如精密检测等多个应用领域发挥作用.
수선응용수학형태학방법진행도상려파,연후대Sobel변연검측산자진행개진,병인입최소이승의합방법진행산법구정,종이구조출일충신적이론체계완비적자상소급변연검측산법,종려파、변연정위、의합구정등다개방면대전통적변연검측산법가이개진,사기가이재제여정밀검측등다개응용영역발휘작용.
In the paper, morphology is used to filter the image firstly, and then the Sobel edge detection operator is improved. Furthermore,the least square fitting method is introduced to polish the edge detection,which compose a new subpixel edge detection algorithm with the complete theoretical system by conjoining morphology, numerical analysing method with edge detection theory. The algorithm improves the performance of classical edge detection methods in filtering,edge locating and fitting method for precision and makes it practical in the accurately detecting field and some other fields