有无噪声对细胞神经网络提取图像边缘的影响

主要运用相关系数，通过对原始图像进行加躁再去噪，计算出用传统的滤波方法如均值滤波、中值滤波和维纳滤波分别对有噪声的图像进行去噪处理后，去噪图像与原始图像的相关系数，验证了去除高斯噪声用一次维纳滤波比较理想；去除椒盐噪声使用中值滤波效果较好。并在第三部分给出了对加噪的图像先进行去噪处理，再用CNN进行边缘提取的结果和对加躁图像先用cNN进行边缘提取再去噪的结果，并将这两种结果与原始图像提取边缘的结果进行比较和观察，得出以下结论：为了避免由于噪声引起的失真对图像真实信息的影响，在进行图像边缘提取之前必须先进行去噪处理再进行边缘提取。

The Impact of Nnoise or Noiseless on Cellular Neural Network to Detect Image Edge

This paper mainly using the correlation index, and through the original image add noise and remove noise calculate the correlation index of the removal-noise image the original image,that using the traditional filtering methods such as average filtering and median filtering and wiener filtering delete the noise of an image . Then confirm this conclusion： The best effect to remove gaussian noise is using wiener filtering and Poemove salt and pepper noise using median filtering will get a better effect.In the third part of this paper puts forward two results ：one is providing the add noise image first remove noise then using CNN to detect the image edge ,the other is providing the noise image first using CNN to detect edge then remove the noise. Via the two results compared with the result of the original image edge detection. Get the following conclusion： in order to avoiding the impact of noise bring to an image , before the image edge detect,we must deal with the noise firstly.