Application of Convolutional Neural Network in Object Recognition of Remote Sensing Image

摘要

Object recognition of remote sensing image is of great theoretical significance and application value in many fields. Faster and more effective object recognition methods are the hot and difficult point in the field of image research. Aiming at the problems of object recognition of remote sensing image, in this paper, the convolutional neural network with inter-class constraint (ICNN) is applied to object recognition of remote sensing image. This method replaces the softmax loss function of traditional convolutional neural network with the inter-loss function to obtain smaller intra-class distance and larger inter-class distance. This method significantly improves the effectiveness of image feature classification. Experiments are conducted on the US Land Use Classification Data Set 21(UCM_LandUse_21), and the experimental results showed that the proposed method can realize the fast and accurate recognition of remote sensing image and has a good promotion significance.