微型电脑应用

卷积是一个常用的图像处理技术,也是研究卷积神经网络的第一步。卷积操作会导致图像变小从而损失图像边界,为了保证卷积后的图像和原始图像尺寸相同,研究了各种边界扩展方法,并在OpenCV中对标准图像进行了各种边界扩展实现以及以高斯-拉普拉斯算子作为卷积核对扩展图像的卷积实现与裁剪。

Convolution is a common image processing technique,and is the first step in the study of convolutional neural network.The convolution operation will cause the image to become smaller and lose the edges of an image.To ensure that the image keeps the size as the original image after convolution operation,various boundary extension methods are studied and realized by standard image in OpenCV,and convolution is realized and cut by Gauss-Laplacian operator.