It is very important to determine the magnetic object location using the magnetic anomaly. This paper presents a method for determined the magnetic object location using the geomagnetic field intensity. The magnetic o...It is very important to determine the magnetic object location using the magnetic anomaly. This paper presents a method for determined the magnetic object location using the geomagnetic field intensity. The magnetic object center position can be obtained by the amplitude of the analytic signal( AS) of geomagnetic total intensity. The vertical distance can be obtained by the value of magnetic anomaly on center position. But the vertical distance from measuring plane to magnetic object is not relatively precious because of not eliminating the geomagnetic normal field in the calculation process. The method of modeling geomagnetic normal field in small space is proposed based on Taylor polynomial. The impact of geomagnetic variation field for the precious of model is analyzed. Through the measured data and the model data,total intensity magnetic anomaly on the measuring plane is obtained which is used for calculating the vertical distance of magnetic object by calculating model. The experimental tests have been conducted on open field on the campus of Harbin Engineering University. The results show that the calculated results and real values match perfectly using this method.展开更多
语义分割是遥感影像分析中的重要技术之一。现有方法(如基于深度卷积神经网络的方法等)虽然在语义分割中取得了显著进展,但往往需要大量训练数据。基于图模型的马尔可夫随机场模型(Markov random field model,MRF)提出了一种不依赖训练...语义分割是遥感影像分析中的重要技术之一。现有方法(如基于深度卷积神经网络的方法等)虽然在语义分割中取得了显著进展,但往往需要大量训练数据。基于图模型的马尔可夫随机场模型(Markov random field model,MRF)提出了一种不依赖训练数据的无监督语义分割思路,可以有效地刻画地物空间关系,并对地物空间分布的统计规律进行建模。但现有的MRF模型方法通常建立在基于像素或对象的单一粒度基元上,难以充分利用影像信息,语义分割效果不佳。针对上述问题,引入交替方向乘子法(alternative direction method of multiplier,ADMM)并将其离散化,提出了一种像素与对象基元协同的MRF模型无监督语义分割方法(MRF-ADMM)。首先构建像素基元和对象基元两个概率图,其中像素基元概率图用于刻画影像的细节信息,保持语义分割的边界;对象基元概率图用于描述较大范围的空间关系,以应对遥感影像地物内部的高异质性,使分割结果中地物内部具有良好的区域完整性。在模型求解过程中,针对像素和对象基元的特点,提出了一种离散化的ADMM方法,并将其用于两种基元类别标记的传递与更新,实现像素基元细节信息和对象基元区域信息的协同优化。高分二号和航拍影像等不同数据库不同类型遥感影像的语义分割实验结果表明,相较于现有的MRF模型,提出的MRF-ADMM方法能有效地协同不同粒度基元的优点,优化语义分割结果。展开更多
基金Sponsored by the National Natural Science Foundation of China(Grant No.61174192)Specialized Research Fund for the Doctoral Program of Higher Education(Grant No.20112304110003)
文摘It is very important to determine the magnetic object location using the magnetic anomaly. This paper presents a method for determined the magnetic object location using the geomagnetic field intensity. The magnetic object center position can be obtained by the amplitude of the analytic signal( AS) of geomagnetic total intensity. The vertical distance can be obtained by the value of magnetic anomaly on center position. But the vertical distance from measuring plane to magnetic object is not relatively precious because of not eliminating the geomagnetic normal field in the calculation process. The method of modeling geomagnetic normal field in small space is proposed based on Taylor polynomial. The impact of geomagnetic variation field for the precious of model is analyzed. Through the measured data and the model data,total intensity magnetic anomaly on the measuring plane is obtained which is used for calculating the vertical distance of magnetic object by calculating model. The experimental tests have been conducted on open field on the campus of Harbin Engineering University. The results show that the calculated results and real values match perfectly using this method.
文摘语义分割是遥感影像分析中的重要技术之一。现有方法(如基于深度卷积神经网络的方法等)虽然在语义分割中取得了显著进展,但往往需要大量训练数据。基于图模型的马尔可夫随机场模型(Markov random field model,MRF)提出了一种不依赖训练数据的无监督语义分割思路,可以有效地刻画地物空间关系,并对地物空间分布的统计规律进行建模。但现有的MRF模型方法通常建立在基于像素或对象的单一粒度基元上,难以充分利用影像信息,语义分割效果不佳。针对上述问题,引入交替方向乘子法(alternative direction method of multiplier,ADMM)并将其离散化,提出了一种像素与对象基元协同的MRF模型无监督语义分割方法(MRF-ADMM)。首先构建像素基元和对象基元两个概率图,其中像素基元概率图用于刻画影像的细节信息,保持语义分割的边界;对象基元概率图用于描述较大范围的空间关系,以应对遥感影像地物内部的高异质性,使分割结果中地物内部具有良好的区域完整性。在模型求解过程中,针对像素和对象基元的特点,提出了一种离散化的ADMM方法,并将其用于两种基元类别标记的传递与更新,实现像素基元细节信息和对象基元区域信息的协同优化。高分二号和航拍影像等不同数据库不同类型遥感影像的语义分割实验结果表明,相较于现有的MRF模型,提出的MRF-ADMM方法能有效地协同不同粒度基元的优点,优化语义分割结果。