• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>Asia-Pacific Conference on Synthetic Aperture Radar >Segmentation of human body active millimeter-wave image based on deep learning
【24h】

Segmentation of human body active millimeter-wave image based on deep learning

机译：基于深度学习的人体主动毫米波图像分割

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

摘要

Millimeter-wave imaging technology is very suitable for detecting weapons and other dangerors things concealed under clothe. So it is becoming more and more popular for security checking at key sites like railway stations and airports. Post processing of the millimeter-wave image is very important for the sucessful application of this technology. A lot of post processing work relate to human body segmentation, however this problem is not treat as general problem before. In this paper, we try to solve the problem base on deep learning. To apply deep learning, we collect and annotate a small size data set of active millimeter-wave images of human body. With the data set, we design and implement models based on deep learning to segment the human body into meaningful regions. The performance of the models are carefully compared and analyzed which guides us to a decent solution for security checking.
机译：毫米波成像技术非常适合于探测隐藏在衣服下的武器和其他危险物。因此，在火车站和飞机场等关键站点进行安全检查变得越来越普遍。毫米波图像的后处理对于成功应用该技术非常重要。许多后处理工作与人体分割有关，但是此问题以前并未被视为一般问题。在本文中，我们尝试解决基于深度学习的问题。为了应用深度学习，我们收集并标注了人体活动毫米波图像的小尺寸数据集。利用数据集，我们基于深度学习来设计和实现模型，以将人体分割成有意义的区域。仔细比较和分析了模型的性能，这为我们提供了一个不错的安全检查解决方案。

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