基于深度学习的遥感影像数据库构建分析  

Research on Construction of Database of Remote-sensing Image Based on Deep Learning

作　　者：吴志祥 汤洁玉 张文泽 WU Zhi-xiang;TANG Jie-yu;ZHANG Wen-ze(School of Geosciences and Info-physics,Central South University,Hunan Changsha 410000 China;Hunan Polytechnic of Water Resources and Electric Power,Hunan Changsha 410000 China)

机构地区：[1]中南大学地球科学与信息物理学院,湖南长沙410083 [2]湖南水利水电职业技术学院,湖南长沙410131

出　　处：《科技创新与生产力》2018年第3期65-67,70,共4页Sci-tech Innovation and Productivity

摘　　要：基于深度学习遥感影像目标识别前提在于构建大规模高质量影像识别基准库,以便模拟,识别算法快速准确地提取图像特征。本文详细分析了深度卷积网络驱动下遥感影像数据库的整体构建方法、规模、属性等信息,同时展望了遥感影像数据构建的趋势。The premise of object of remote-sensing image based on deep learning is construction of reference library of large-scale and high quality image recognition in order to quickly and precisely extracts image features by mode and algorithm.In this paper,the authors expounded integral construction methods,scale and property under condition of internet-driven of deep convolution,and prospected trend of building data of remote-sensing image.

关 键 词：深度学习 遥感 影像数据库 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]