尺寸测量中的边缘检测算法研究

主要介绍了一种改进的边缘检测算法及相应计算模板和公式，并对算法的误差进行了分析。因为传统的边缘检测算法在理论上较成熟，但在实际应用时可操作性较差。所以针对零件图像测量的实际应用情况，使用基于Sobel算子的改进的方向算子，对灰度图像进行处理得到梯度图像。综合应用其他算法，实现了对目标边缘的准确检测。然后在梯度图像上沿目标边缘的梯度方向进行多项式插值法亚像素细分计算，对目标边缘进行亚像素精确定位。在文章的最后，用实例说明了本算法的可行性。

Study of Edge Detection in Size Measurement

A improved edge detecting algorithm, mask, and calculating formula are introduced, and the cause of error of the operator and derives the error formula are analyzed.The traditional algorithm has not favourable maneuverability used in actual application. So,the new edge detecting algorithm for the imaging measurement use the direction mask based on classical Sobel mask in converting gray into grads image,and apply the polynomial interpolation to subpixel edge detecting at grads direction of edge on grads image. Finally,some examples are given to demonstrate the effectiveness of the method.