结合小波去噪的改进型Canny边缘检测算法

Improved Canny Edge Detection Algorithm Combined with Wavelet Denoising

针对传统Canny边缘检测算法采用高斯滤波器对图像进行平滑去噪时容易引起去噪图像过度光滑的问题,提出了一种将小波去噪引入到Canny算法中的改进算法。使用小波滤波方法可以较好地去除图像噪声,通过计算梯度算子的方向和幅值,采用非极大值抑制与双阈值方法来连接边缘。实验结果表明:改进后的Canny算法不仅对噪声具有较高的抑制能力,而且还使图像边缘轮廓更加清晰。

For the traditional Canny edge detection algorithm in which using the Gauss filter to smooth the im- age denoising is easy to cause the problem of excessive smooth noise, this paper presents a wavelet denoising algorithm based on Canny algorithm. Using the Wavelet filtering method can better remove the noise in the image, according to the calculation of amplitude and direction of the gradient operator, non -maxima suppression method and double threshold are adopted to connect the edge. The experimental results show that the improved Canny algorithm not only has higher inhibition ability for noise, but also makes image profile more clear.