In view of the fact that the deep convolutional neural network has stronger ability of feature learning and feature expression, an exploratory research is done on feature extraction and classification for high resolution remote sensing images. Taking the Google image with 0.3 meter spatial resolution in Ludian area of Yunnan Province as an example, the image segmentation object was taken as the basic unit, and the pre-trained AlexNet deep convolution neural network model was used for feature extraction. And the spectral features, AlexNet features and GLCM texture features are combined with multi-kernel learning and SVM classifier, finally the classification results were compared and analyzed. The results show that the deep convolution neural network can extract more accurate remote sensing image features, and significantly improve the overall accuracy of classification, and provide a reference value for earthquake disaster investigation and remote sensing disaster evaluation.

ABSTRACT:The rapid collection of Unmanned Aerial Vehicle (UAV) remote sensing images plays an important role in the fast submitting disaster information and the monitored serious damaged objects after the earthquake. However, for hundreds of UAV images collected in one flight sortie, the traditional data processing methods are image stitching and three-dimensional reconstruction, which take one to several hours, and affect the speed of disaster response. If the manual searching method is employed, we will spend much more time to select the images and the find images do not have spatial reference. Therefore, a near-real-time rapid georeference method for UAV remote sensing disaster data is proposed in this paper. The UAV images are achieved georeference combined with the position and attitude data collected by UAV flight control system, and the georeferenced data is organized by means of world file which is developed by ESRI. The C # language is adopted to compile the UAV images rapid georeference software, combined with Geospatial Data Abstraction Library (GDAL). The result shows that it can realize rapid georeference of remote sensing disaster images for up to one thousand UAV images within one minute, and meets the demand of rapid disaster response, which is of great value in disaster emergency application.