基于启发式图搜索的遥感影像道路半自动提取

启发式图搜索法用于线状目标识别的原理是：用图结构表示边缘点和边缘段，根据启发函数计算顶点权值，在图的路径上建立相应的代价函数，通过在图中搜索对应的最小代价的通道以找到最优路径。图搜索法是一种全局最优方法，它在受噪声影响较大时效果仍然较好。文中使用了启发式图搜索法（A*算法）实现了道路的半自动跟踪。它的基本思路是：首先利用自适应平滑滤波算子进行道路信息增强，然后对传统的道路数学模型进行了进一步的扩展，突出了对道路几何特性和辐射特性的描述，并依此构建图搜索的代价函数，实现了基于启发式图搜索法A*算法的道路半自动跟踪。经实验证明，该方法进行遥感影像的道路半自动提取效果较好。     

通海地磁台(L=1.03)Pc3地磁脉动的出现率

利用通海地磁台(L=1．03)1999年SMALL磁通门磁力仪记录资料，对Pc3脉动出现率的统计结果表明，日出现频次有两个峰值，即清晨06：00-07：00LT(地方时)出现频次最高，午后14：00～15：00LT为次峰值。Pc3脉动月出现频率最高的是4月和12月，其次是8月和9月；最少的是6月和7月。按季节统计结果为春秋季出现频次最高，冬季次之，夏季最低。     

Confidence Evaluation of Medium-and Long-Term Earthquake Prediction by Use of the Analytical Hierarchy Process

This paper offers an approach to the relative confidence level of medium- and long-term earthquake prediction methods by use of the analytical hierarchy process (AHP). We construct the criterion judgment matrices according to the results of an expert questionnaire known as the Delphi scheme. From the matrices we can obtain a scale for the relative preference or priority weight of each factor. The quantitative weights might be applied to make synthetic prediction of earthquake risk areas in the northern part of the North China seismic region.     

Research on seismic intensity zonation by use of the response intensity of historical earth－quakes───with the central part of Shanxi Province as an example

Ｒｅｓｅａｒｃｈｏｎｓｅｉｓｍｉｃｉｎｔｅｎｓｉｔｙｚｏｎａｔｉｏｎｂｙｕｓｅｏｆｔｈｅｒｅｓｐｏｎｓｅｉｎｔｅｎｓｉｔｙｏｆｈｉｓｔｏｒｉｃａｌｅａｒｔｈ┐ｑｕａｋｅｓｗｉｔｈｔｈｅｃｅｎｔｒａｌｐａｒｔｏｆＳｈａｎｘｉＰｒｏｖｉｎｃｅａｓａｎｅｘ...     

淮阴台井下测震震级偏差初探

依据台站两年来实测的地震震级和江苏省区域测震台网定的震级进行比较,计算出各个地震的偏差值和台站的震级校正值,并获得相应的标准误差,最后对影响震级差值因素进行了探讨。     

利用普通收音机在短周期测震记录仪上打授时台秒号的简单接口

详细叙述了利用普通收音机在短周期测震记录仪上打授时台秒号的接口,它的设计原理、安装与使用。经实践、成本不高,性能确可满足需要。     

减小水平分量磁变仪记录非线性的技术措施

讨论了减小水平分量磁变仪记录非线性的技术措施。选择适宜的标度值(C/M)、选取适当的入射角α_0、选择适当的记录距离R是改善水平分量磁变仪记录状态的主要途径。文中对比作了具体分析。     

Prediction of the Earlier Strong Aftershocks in the Earthquake Sequences by the Pseudo-Periodicity Method

We propose the pseudo-periodicity method and its quantitative prediction indexes for the occurrence time of earlier strong aftershock. We conducted tests of regressive prediction, and the R-value of the tests is 0.45, indicating that this method is effective for prediction.