Deep Learning-Based Classification of Remote Sensing Image

摘要

Deep Learning networks have sharply increased over the past 10 years, and deep Learning-Based Classification of Remote Sensing Image has attracted extensive interest. We trained a multilayer deep learning network to classify the 8 thousand unlabeled remote sensing images from Internet into the 600 different classes. In order to improve the efficiency, and shorten the experiment time, we also used pre-trained model, NVIDIA GTX970 GPU, 32GB internal memory, 10T hard-disk. We achieved error rates of 9.7% which work went relatively well than the traditional machine learning techniques. Deep learning-based network can achieve the classification of unlabeled data without any manual intervention. Compared to those usual machine learning algorithm, accuracy and speed of deep learning-based classification network is more faster and accurately.