中文摘要：

为了提高边缘检测精确度与抗噪性能，该文提出一种基于保角相位的图像边缘检测新方法。该方法首先利用保角单演信号能够表达不同本征维数的图像局部结构的特点，采用指数函数计算相位偏差，有效地抑制了相位一致模型边缘检测中产生的伪边缘和噪声，提高了边缘检测的精确度；其次，利用Poisson核在空域中有解析表示的优势，降低了算法复杂度。仿真实验结果表明，与现有的相位一致性图像边缘检测方法相比，该方法提取的图像边缘更精确、更完整、更均匀，对噪声具有更好的鲁棒性，同时，计算复杂度较低。

英文摘要：

To improve the image edge detection accuracy and anti-noise performance, a new approach for image edge detection based on conformal phase is proposed. Firstly, the proposed approach can effectively improve the precision of edge detection and restrain the false edge and noise by using respectively the conformal monogenic signal which could express local structure of the image with different intrinsic dimensions and an exponential function to calculate the phase deviation. Secondly, it can reduce the complexity of the algorithm by taking advantage of the Poisson kernel of existence of analytic representation in spatial domain. To demonstrate the advantages, the proposed approach is compared with the existing methods of phase congruency based edge detection. The simulation experiment results show that the proposed approach can extract image edge more accurately, more completely, and more uniformly, with better robustness to noise and lower computational complexity.