• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Journal of Medical Imaging and Health Informatics >3D Shape Descriptor for Activities of Daily Living (ADLs) Recognition Based on Kinect-Like Depth Images
【24h】

3D Shape Descriptor for Activities of Daily Living (ADLs) Recognition Based on Kinect-Like Depth Images

机译：基于Kinect深度图像的用于日常生活活动（ADL）识别的3D形状描述符

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

The monitoring of human activities of daily living (ADLs) is one of the major indirect assessments for obtaining the functional status of a person suffering with cognitive and physical impairment. The automated human activities monitoring from video has been widely investigated in the computer vision community. However, most of the human activities monitoring systems were based on conventional cameras which provided colour (RGB) images. This paper reports on the feasibility of general 3D shape descriptors in recognizing the dynamic human activity sequence based on Kinect-like depth images. In this study, general 3D shape descriptors; (1) shape distribution; (2) local spin image; (3) global spin image; and (4) shape histogram were compared using several proposed similarity measurement functions and performance frameworks. It was found that the shape distribution (with AUC-ROC = 0.6181) and local spin image (with AUC-ROC = 0.6172) considerably outperformed the rest of the 3D shape descriptors.
机译：监视人类的日常生活（ADL）是获得具有认知和身体障碍的人的功能状态的主要间接评估之一。通过视频进行自动人类活动监控已在计算机视觉社区中得到广泛研究。但是，大多数人类活动监视系统都是基于提供彩色（RGB）图像的常规相机。本文报道了一般的3D形状描述符在基于类似Kinect的深度图像识别动态人类活动序列中的可行性。在这项研究中，通用的3D形状描述符; （1）形状分布; （2）局部自旋图像; （3）全局自旋图像; （4）使用几种拟议的相似性度量功能和性能框架对形状直方图进行了比较。发现形状分布（使用AUC-ROC = 0.6181）和局部旋转图像（使用AUC-ROC = 0.6172）明显优于其余3D形状描述符。

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