• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>International conference on signal and information processing, networking and computers >Research on Target Recognition Technology of Satellite Remote Sensing Image Based on Neural Network
【24h】

Research on Target Recognition Technology of Satellite Remote Sensing Image Based on Neural Network

机译：基于神经网络的卫星遥感图像目标识别技术研究

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

With the rapid development of satellite remote sensing technology, the resolution of satellite image is getting higher and higher, and more and more satellite data can be obtained on the ground. Traditional artificial image translation methods can not deal with massive data, and can not efficiently, quickly and accurately obtain the information of interested objects. In view of this problem, considering that the depth convolution neural network technology has achieved good results in the natural image target recognition, this paper uses the typical depth neural frame Faster R-CNN as the basic frame, and uses the image augmentation method to enhance the accuracy and generalization ability of the neural network model, and multi-resolution optical remote sensing image data to achieve automatic target recognition processing. The results show that the proposed method can translate images automatically and quickly, the recognition rate of ship and other targets is better than 75%.
机译：随着卫星遥感技术的快速发展，卫星图像的分辨率越来越高，更高，并且在地面上可以获得越来越多的卫星数据。传统的人工形象翻译方法无法处理大规模数据，无法有效，快速准确地获取感兴趣的对象的信息。鉴于此问题，考虑到深度卷积神经网络技术在自然图像目标识别中取得了良好的结果，本文使用典型的深度神经框架更快的R-CNN作为基本帧，并使用图像增强方法来增强神经网络模型的准确性和泛化能力，以及多分辨率光学遥感图像数据实现自动目标识别处理。结果表明，该方法可以自动快速翻译图像，船舶和其他目标的识别率优于75％。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号