A combination of improvements based on Sobel and polynomial interpolation Silicon sheet steel image sub-pixel edge detection algorithm is introduced in this paper. The two-direction template is added to the classical Sobel algorithm in this algorithm and on the edge of the gradient image with a reasonable weight and threshold method is used to get an accurate single-pixel edge. Polynomial interpolation of the direction of the gradient on the edge of the single-pixel is calculated to get a more precise sub-pixel edge location. This detection is accurate, which can be used to detect the Silicon sheet steel image sub-pixel edge on the assembly line.%提出一种基于Sobel和多项式插值法结合的改进的黑片图像亚像素边缘检测算法.算法在经典Sobel算法的基础上增加了两个方向模板,并使用对边缘梯度图像加上合理的权值和阈值的方法得到准确的单像素边缘.在单像素梯度边缘上进行有方向的多项式插值计算得到更精确的亚像素边缘定位.该方法的检测精度高,能够应用于黑片的在线检测.
展开▼
