Classification of Landscape Pictures Based on Residual Network

The continuous development of artificial intelligence has brought great convenience to our life and work. One of the more mature applications in artificial intelligence is the function of image classification, which provides a new method for us to process massive data. When faced with an explosion of image information on the Internet, we can use deep learning to categorize images and select only the categories we want to stay in. This will save us a lot of time and reduce our workload. This paper is based on deep learning to classify several common types of landscape maps, so that we can easily choose the scenery we want to visit. The residual network structure is adopted in this paper, and the accuracy can reach 87.6% only after low rounds of training, which is better than the previous shallow network structure.