Abstract: For image measurement, the precise positioning and rapid detection of the image edge are critical factors. In order to realize the rapid and high precision edge positioning and detection, a novel method is proposed based on an improved Zernike method. According to the geometric information of the target, the primary processing of the image is operated based on four-direction-template Sobel operators, then sub-pixel edge positioning can be obtained using only one of moment-templates. Experimental results demonstrate that the proposed method has higher precision, and less running time a quarter than that based on the original method. The novel method has a good processing efficiency, which can be used to position and detect the image edge with the geometric characteristics, with sub-pixel positioning precision and real-time detection advantages, and has a broad application prospect.

Key words: Sobel operator, Zernike moment, sub-pixel, edge detection

摘要： 在图像测量中，图像边缘的精确定位与检测是影响测量精度的关键。为了实现快速、高精度的图像边缘定位与检测，提出了一种改进的Zernike方法，采用四个方向模版Sobel算子对图像初处理，利用被测物的几何信息，只使用零阶矩实现对边缘的亚像素定位。实验结果表明改进算法比原算法具有更高的精度，而运行的时间不到原算法的1/4。该算法具有良好的处理效率，能够满足一般具有几何特征图像边缘实时、亚像素精度定位与检测的要求，具有广阔的应用前景。

关键词: Sobel算子, Zernike矩, 亚像素, 边缘检测