基于高斯曲面拟合的亚像素边缘检测算法

Sub-pixel Edge Detection Algorithm Based on Gauss Surface Fitting

提出了一种基于高斯曲面拟合的亚像素边缘检测算法。该算法首先使用传统Canny算子进行边缘像素级的粗定位,再以粗定位边缘像素为中心取5×5窗口获取梯度幅值,对其进行高斯曲面拟合,在梯度方向上寻找局部极值点,从而达到边缘亚像素级的精细定位的效果。文中介绍了算法的理论依据,给出推导过程,并用螺旋测微器的标准间隙进行实例分析,对结果进行数据分析并与其他方法进行了比对。验证算法的分辨率为0.02个像素,精度达到0.1个像素,有效地提高了图像边缘提取的精确度。

A sub-pixel edge detection algorithm with 0.02 pixels resolution and O. 1 pixels accuracy based on gauss surface fitting technique is proposed. The coarse-fine algorithm consists of two steps. Firstly, the traditional Canny operator is used to detect the coarse position of an edge .Then, the sub-pixel edge position is determined by the gauss surface fitting around the center of the corresponding coarse position with 5 ×5 pixels window. The sub-pixel edge position is the local peak of the intersection curve between the vertical plane in the gradient direction of the edge and the fitted gauss surface. In the paper, the derivation of the sub-pixel edge detection algorithm is presented. The sub-pixel accuracy is verified by the measurement with standard clearance of micrometer caliper and compared with other traditional subpixel edge detection algorithms.