对永胜6级地震的预报及依据

2001年10月27日在云南永胜发生6.0级地震，震前四川省地震局有较准确的中期和短临预报，加强了川滇交界地区的地震监测预报工作，并且向四川省政府作了汇报，取得了明显的社会效益。经验性地震预报重视川滇菱形地块西界断裂带动态活动图像的分析，尤其对历史强震破裂空段附近出现的中小震活动带或新活跃区段的研究；同时，重视4级前震群、川滇交界地区地下水位、水温、形变观测异常的跟踪分析和震情预测。     

基于树结构的数字遥感图像存储方法与应用

数字遥感图像一般是采用栅格结构来存储。图像中地物的结构特征是隐含其中的。该文从图像识别角度，重点研究了基于树结构的数字遥感图像存储方法。这种方法与遥感中常用的四又树等存储方法不同，它注重图像中地物本身的结构特征。通过树结构使像元按地物目标形成一个个的组合体。该文以机场跑道识别为例，介绍了基于树结构的数字遥感图像存储方法的应用以及将图像变为树的算法。     

三种地统计学图像纹理用于遥感图像分类的比较

该文运用LANDSAT TM数据对三种地统计学纹理量测方法用于图像分类的性能进行了比较和分析。研究发现，基于绝对值变差函数的纹理具有更好的性能。经典的变差函数和基于方根的变差函数的性能依赖于提取纹理时所用的窗口大小。     

策勒绿洲植被覆盖动态变化遥感研究

使用经严格配准的2个时相的TM和HRV图像数据为基础底料编制规一化植被指数图（NDVI），根据策勒绿洲特有的生态环境特征，研究植被指数灰度级与盖度级的对应函数关系，将其转换成植被盖度图。最终利用植被盖度图像所提供的各盖度级的数量和空间分配状况来评价图像所包括时段，即1990年至1998年期间植被环境质量的变化，为该区域生态环境综合治理提供新的研究方法。     

图像处理技术在地质灾害监测中的应用

三维激光微位移监测技术是一种对地质灾害进行长期监测的新方法。该技术用激光作为光源，用CCD摄像器作为镜头，通过图象采集卡把视频信号采入计算机中，用与该仪器配套的专门软件对采集到的光斑图像进行处理，计算出光斑的三维中心坐标值，把该值和原点坐标值进行比较，就可以算出灾害体滑动的距离。该系统可以对灾害体进行长期、非接触监测，监测时间、监测频率可以任意设定。文章概括介绍该监测的原理，各个组成部分，重点介绍软件部分，即采集的视频信号进入计算机后的图像处理过程和数据库结构。该监测系统已在三块地区试运行了3个月，效果较好。     

声波扫描成像在灾害防治及环境保护中的应用

利用超声波扫描成像测井技术，在地质灾害的预防与环境保护中可以检测由于采矿而形成的冒落带、裂隙带、离层带、洞穴及软弱层的存在状况及部位，测定基岩风化蚀变程度及深度，在地质灾害与环境保护的治理时可以监测治理的质量，提供了直观精确可靠的基础资料。     

Morphological analysis of active Mount Nemrut stratovolcano, eastern Turkey: evidences and possible impact areas of future eruption

Mount Nemrut, an active stratovolcano in eastern Turkey, is a great danger for its vicinity. The volcano possesses a summit caldera which cuts the volcano into two stages, i.e. pre- and post-caldera. Wisps of smoke and hot springs are to be found within the caldera. Although the last recorded volcanic activity is known to have been in 1441, we consider here that the last eruption of Nemrut occurred more recently, probably just before 1597. The present active tectonic regime, historical eruptions, occurrence of mantle-derived magmatic gases and the fumarole and hot spring activities on the caldera floor make Nemrut Volcano a real danger for its vicinity. According to the volcanological past of Nemrut, the styles of expected eruptions are well-focused on two types: (1) occurrence of water within the caldera leads to phreatomagmatic (highly energetic) eruptions, subsequently followed by lava extrusions, and (2) effusions–extrusions (non-explosive or weakly energetic eruptions) on the flanks from fissures. To predict the impact area of future eruptions, a series of morphological analyses based on field observations, Digital Elevation Model and satellite images were realized. Twenty-two valleys (main transport pathways) were classified according to their importance, and the physical parameters related to the valleys were determined. The slope values in each point of the flanks and the Heim parameters H/L were calculated. In the light of morphological analysis the possible impact areas around the volcano and danger zones were proposed. The possible transport pathways of the products of expected volcanic events are unified in three main directions: Bitlis, Guroymak, Tatvan and Ahlat cities, the about 135 000 inhabitants of which could be threatened by future eruptions of this poorly known and unsurveyed volcano.     

The role of sediment specific gravity and availability on cluster evolution

The objective of this experimental study was to account for the role of sediment availability and specific gravity on cluster formation and cluster geometric characteristics (spacing and size). To isolate the effects of sediment availability and specific gravity on cluster evolution, mono‐sized spheres were used to simulate the cluster evolutionary cycle. Overall, twelve experimental runs were carried out in the laboratory flume. Six of these tests were performed by using glass spheres (specific gravity, SG = 2·58) and the other six by employing an equal combination of glass and Teflon spheres (SG = 2·12) of the same diameter to evaluate the role of specific gravity on cluster evolution. The three sediment availability conditions that were investigated here simulated isolated gravel elements, pool–riffle sequences and densely packed gravel‐bed. An advanced image analysis technique was employed to track the evolution of cluster microforms and provide quantitative information about the size and shape of clusters and the number of clusters per unit area. The results of this study showed that: (1) sediment availability affects the architecture and size of cluster microforms; and (2) clusters consisting of mono‐sized sediments start disintegrating at twice the incipient conditions. By performing complementary tests for the isolated gravel elements case, it was found that the evolutionary cycle of individual clusters could be described as follows, in order of increasing stress: no cluster→two particle cluster→comet→triangle→rhomboid→break up. Copyright © 2003 John Wiley & Sons, Ltd.     

遗传算法与单纯形法组合的影像纹理分类方法

提出遗传算法(简称GA)与单纯形法组合的影像纹理分类方法(简称GASPX)。单纯形法是一种局部搜索方法，它通过反射，扩张，收缩操作，求得新的单纯形点，组成新单纯形，新单纯形比前一个单纯形更接近局部最优解。这种寻优方法收敛速度快，它与GA组合起来可以改善单独使用GA收敛速度慢的缺陷。由于在组合算法中是多个单纯形的局部区域的并行搜索，避免GA优化过程中过早收敛于局部最优解的现象出现。通过5种不同类别航空影像纹理识别的试验，并与GA的结果作对比，结果表明GASPX法优于GA法。     

影像匹配中几种相似性测度的分析

对基于灰度影像匹配中的几种相似性测度进行了研究，分析了每种测度的直观意义、限制条件以及改进措施，最后通过实验结果验证了上述结论。