基于Zernike图像矩的理想边缘模型,利用Zernike矩定义及其旋转不变特性,提出一种新的基于4阶方向角的Zernike矩亚像素边缘检测算子;为了提高边缘算子定位速度,首先基于9×9尺寸模板对Zernike图像矩0到4阶正交复数多项式进行了计算,推导出基于4阶方向角的边缘检测算子参数模型;最后将边缘算子应用在理想图像与实际图像上,检测结果表明,相比于目前提出的算法,基于4阶方向角的边缘检测算子具有更高的检测精度。
Based on the Zernike image moments edge model, provides a new edge operator method based on 4 orders direction angle by u-sing the Zernike moments definition and its rotation invariance. In order to improve the edge operator positioning speed, the orthogonal com- plex polynomial from 0 order to 4 orders of the Zernike moments based on 9 × 9 mask coefficient was given, the edge operator parameter mod-el was deduced. Finally, the edge operator is applied to the ideal binary image and a real image, the edge detection results showed that the 4 orders direction angle model has a higher accuracy compared to the existed methods.