翻译：卷积层是深度卷积神经网络中最重要的组成部分主要作用是提取输入数据的特征。在卷积层CNN通过大小相同的卷积核按照一定的步长stride在输入的图层上滑动计算每个位置的元素乘积之和并将结果保存到输出特征图中。

The convolutional layer is the most important component of deep convolutional neural networks, mainly used to extract features from input data. In the convolutional layer, CNN slides a convolution kernel of the same size on the input layer with a certain stride, calculates the sum of the element-wise multiplication at each position, and saves the result to the output feature map.