期刊文献+
共找到2篇文章
< 1 >
每页显示 20 50 100
Use of SAR interferometry for monitoring illegal mining activities: A case study at Xishimen Iron Ore Mine 被引量:7
1
作者 Ji Maowei Li Xiaojing +2 位作者 Wu Shunchuan Gao Yongtao Ge Linlin 《Mining Science and Technology》 EI CAS 2011年第6期781-786,共6页
The development and application of the ''digital mine'' concept in China depends heavily upon the use of remote sensing data as well as domestic expertise and awareness. Illegal mining of mineral resou... The development and application of the ''digital mine'' concept in China depends heavily upon the use of remote sensing data as well as domestic expertise and awareness. Illegal mining of mineral resources has been a serious long term problem frustrating the Xishimen Iron Ore Mine management. This mine is located in Wu'an county in Hebei province, China. Illegal activities have led to enormous economic losses by interfering with the normal operation of the Xishimen mine and have ruined the surrounding environ- ment and the stability of the Mahe riverbed the crosses the mined area. This paper is based on field recon- naissance taken over many years around the mine area. The ground survey data are integrated with Differential Synthetic Aperture Radar Interferometry (D-InSAR) results from ALOS/PALSAR data to pin- point mining locations. By investigating the relationship between the resulting interferometric deforma- tion pattern and the mining schedule, which is known a priori, areas affected by illegal mining activities are identified. To some extent these areas indicate the location of the illegal site. The results clearly dem- onstrate D-InSAR's ability to cost-effectively monitor illegal mining activities. 展开更多
关键词 D-InSAR Monitoring illegal mines Surface deformation
下载PDF
基于表观模型的人脸特征点提取 被引量:4
2
作者 闫保中 何伟 韩旭东 《应用科技》 CAS 2020年第6期47-52,共6页
针对在人脸特征点提取过程中模型预估能力不足,拟合过程中的非法形变导致的特征点提取速度慢和提取精度不高的问题,本文提出了一种基于主动表观模型(AAM)的人脸特征点提取算法的改进算法。将头部作初始分类,不同的头部姿态选取不同人脸... 针对在人脸特征点提取过程中模型预估能力不足,拟合过程中的非法形变导致的特征点提取速度慢和提取精度不高的问题,本文提出了一种基于主动表观模型(AAM)的人脸特征点提取算法的改进算法。将头部作初始分类,不同的头部姿态选取不同人脸模型进行拟合,这样能避免初始化模型与真实特征点位置相差过大,从而使得模型能更快的收敛,提高特征点提取速度。同时提出一种方法,对拟合过程中的形状变量加以限制,能有效过滤掉不满足人脸形状的特征点模型,防止拟合过程中的非法形变,使提取的人脸特征点更接近真实位置。实验结果表明,改进的基于AAM的人脸特征点提取算法在时间效率和准确率上都所有提高。 展开更多
关键词 表观模型 特征提取 人脸特征点 主动表观模型 非法变形 方向梯度直方图特征 拟合 头部姿态
下载PDF
上一页 1 下一页 到第
使用帮助 返回顶部