中文摘要：

为提高图像弱边缘的检测效果，解决图像细节边缘的检出问题，最大限度地保留图像内容信息，提出了一种通过对图像进行四相位移动模糊差分，利用图像边缘灰度梯度变化的相关性，来全面的检测图像内部强、弱边缘的图像边缘检测新算法。该算法可有效的检测图像细节等弱边缘，分割滤除背景，完整的保持图像内目标的结构特征，同时对于图像质量要求较低，具有较高的边缘检测特性。

英文摘要：

Edge detection is a key problem in image-analysis processing. Aiming at detect the detailed edges, i.e. all the sharp and blur edge, a new algorithm is proposed based on fuzzy moving difference of four-phase which infers from the relativity of the image grey changes. Experimental result shows that the algorithm is able to detect the detailed edge effectively and preserve the accordingly structured features from the background. And the algorithm even processes a high performance with the poor quality image.