本文引入小波多尺度边缘检测,将小波的尺度因子与不同方向的梯度联系起来在梯度大的方向(垂直边缘方向)采用小尺度的因子,突出细节.在梯度小的方向(沿着边缘方向)采用大尺度因子,可提高信噪比.试验结果表明:该算法具有很好的边缘细节的保持性,较常见的P-M模型滤波方法算法效果有明显优势.
This paper introduces the wavelet multi--scale edge detection, which correlated wavelet scale factor and different direction gradient. Using small scale factor in the large gradient direction , which can highlight the details. Using a large scale factor in the small gradient direction, which can improve signal to noise ratio. The test results show that., the algorithm has very good edge retention, has obvious advantages compared with P--M model filtering algorithm.