现有的基于小波变换的图像边缘提取方法会导致边缘细节的损失且边缘位置会发生偏移，因此本文给出了一种改进的边缘检测方法。该算法先对图像进行平滑处理，然后用小波变换提取边缘。传统的平滑方法避开了边缘的方向性，且对图像的边缘保持效果不佳。本文提出了基于边缘方向性的平滑算法，该算法在处理边缘像素时可自动搜索边缘方向进行平滑，用该算法和小波方法结合进行边缘检测。仿真实验给出的实验结果有力地证明了该方法的有效性。

The existing image edge detection method based on wavelet transform may lose some details of edge and have the drift of the edge position across different scales,under this situation,we present an improved method for edge detection based on wavelet transform.This method smoothes the image first and then detects the image edge.Traditional image smoothing methods avoid the direction of edge and can not retain the image edge efficiently.This paper presents a new image smoothing algorithm based on edge direction.The smoothing algorithm smoothes the image on the direction of edge,and then detects the image edge with this algorithm and wavelet transform.Finally,the experimental results show that the proposed algorithm is very reliable and efficient.