一种基于多项式插值改进的亚像素细分算法

A Improved Subpixel Edge Detecting Algorithm Based on Polynomial Interpolation

摘要: 介绍了一种基于多项式插值法的改进的亚像素细分算法及相应计算模板和公式,并对算法的误差进行了分析.本算法应用在经典Sobel算子基础上构造出的方向模板,对灰度图像进行处理,得到梯度图像,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.实例说明本算法是可行的.

Abstract: A improved subpixel edge detecting algorithm based on polynomial interpolation, masks and calculating formulae is introduced, the error cause of the operator is analyzed, and the error formula is derived. The algorithm uses direction masks based on classical Sobel masks in converting gray into gradient images and applies the polynomial interpolation to subpixel edge detecting at the gradient direction of edge on gradient images. Some examples are presented to demonstrate the effectiveness of the method.