多视纹理映射中无需设定偏差的快速可见性检测

在多视纹理映射中,Z缓冲区算法和光线追踪算法是两种主流的可见性检测算法。但其精度受限于人工设定的偏差。本文提出一种无需设定偏差的可见性检测方案。首先,根据倾斜摄影的投影参数设计了一种基于着色器编程的影像模拟算法,生成各影像视角下的初始可见性图(initial visibility map,IVM)。其中,完全不可见面元在模拟的过程中会被深度测试自动剔除。然后,基于矢量栅格化准则和像素深度对IVM中的可见面元执行投影区域优化(projection coverage refinement,PCR)。最后,提出惰性投影(lazy projection,LP)和迭代点边采样(iterative vertex-edge sampling,IVES)快速区分出IVM中的完全可见面元和非完全可见面元。利用两个数据集对本文算法进行了论证。结果表明本文算法的效率和精度均优于主流算法。     

大规模地形实时绘制中的地平线遮挡剔除算法

大规模地形场景包含大量的几何数据,无法一次性载入内存,并具有极高的复杂度,因而无法进行实时绘制。本文提出一种用于大规模地形场景的实时地平线遮挡剔除算法。该算法采用四叉树结构构造地形的多分辨率层次细节模型,按照大致由前向后的顺序遍历地形时构建遮挡地平线,应用地形包围盒技术,把地形区域与地平线相比较,以确定何时它们对于视点来说完全被遮挡并剔除这些被遮挡的区域。实验结果表明在贴地漫游的情况下,显著提高了场景绘制速度,证明了此算法的有效性。     

真正射影像生成的多边形反演成像遮蔽检测方法

遮蔽检测是真正射影像生成的关键技术。提出一种基于多边形反演成像(polygon based inversion imaging,PBI)的遮蔽检测方法。利用建筑物表面多边形内部互不遮蔽的特点,以多边形为单元将建筑物逆投影到像方,反演成像时的目标状态,获得目标之间、多边形之间的遮蔽关系。为确保算法的稳健性和保真度,提出:①可疑区域增长法,稳健地栅格化复杂3维建筑物模型;②综合滤波模型用于消除多边形边界噪声。最后利用实例比较z-buffer方法,基于射线角度方法和PBI方法的遮蔽检测效果。结果表明,PBI算法的有效性和稳健性较好。     

Complex correspondences in straight line stereo matching

Partial occlusion and fragmented lines will result in the various cases of straight line correspondences, such as one-to-one, one-to-many or many-to-many ones. However, the complex correspondences, such as one-to-many and many-to-many ones, are usually ignored or cannot be established completely in the existing methods. Here, the essence of the complex correspondences will be analyzed. Based on the two characteristics of a straight line, which are introduced by regarding a straight line as a set of collinear points, the compatibility between the complex correspondences and the uniqueness constraint of point correspondence is proved and a new uniqueness constraint of correspondence for matching lines is proposed. Based on the analysis of the complex correspondences, a new concept of line feature group is defined to describe a set of integral correspondences among straight line features from different images and then a new algorithm for establishing all the correspondences completely is described simply. The experimental results with real stereo images illustrate that the complex correspondences among straight lines are actual cases and can be established effectively.     

基于图割算法的倾斜影像纹理映射优化方法

针对基于倾斜摄影影像重建三维城市模型纹理映射过程中出现的遮挡和纹理碎片化问题,本文提出一种利用摄影光线与模型相交判断遮挡,然后利用图割算法优化纹理的方法。首先,连接摄影中心与三角形面片各顶点,判断连接线段与其他三角形面片的相交情况,从而确定是否遮挡,以选取当前三角形面片对应的未被遮挡影像。然后,根据对应关系,构建能量函数,通过图割算法最小化能量函数以确定最优纹理,使相邻三角形面片对应影像的选择趋于一致,从而解决纹理碎片化问题。通过实验证明,本文方法在大规模三维模型的纹理映射中,能够有效解决映射的遮挡重影现象,并能够极大地避免纹理碎片化问题。     

融合局部形变模型的鲁棒性人脸识别

将测试图像看成是人脸库的线性组合,并用形变模型表示,优化匹配求解组合系数,根据系数的稀疏性进行分类识别。为了进一步提高算法的鲁棒性,采用了分片加权的策略。在公用的人脸数据库上进行广泛的实验,结果表明,平均识别率达到97%以上,在遮挡30%时其识别率仍达到95%以上。本方法对人脸识别问题非常有效,且可以显著提高对伪装、遮挡变化的鲁棒性和稳定性。     

利用LiDAR点云的真正射影像遮蔽检测

针对LiDAR系统真正射影像生成过程中由于高地物投影差产生的遮蔽问题,结合LiDAR点云数据提出了一种新的检测方法,将传统的基于格网的遮蔽区域检测转化为基于多尺度的TIN三角面片遮蔽检测。利用荆州市区获取的LiDAR数据对该方法与传统方法进行了比较,实验证明,该算法通过多重检测后忽略平坦地区,能够快速、准确地检测出遮挡区域。     

移动高程平面约束的多视影像特征点匹配

针对影像匹配中由于重复纹理和遮挡导致的不确定性问题,提出移动高程平面约束的多视影像特征点匹配方法。该方法将多视影像上提取的特征点投影到不同高程的物方格网平面,利用平面上格网单元来约束候选同名特征点,对多视影像进行选择性匹配,避免部分影像遮挡对匹配的影响。实验选取UCX(UltraCamX)数字航空影像进行匹配,取得可靠的匹配结果,从而验证了本文方法的有效性。     

Approximate deflection of rigidly framed earth retaining structures due to an unknown earth pressure distribution

A simple and reliable method for predicting the relationship between lateral displacement and earth pressure for rigidly framed earth retaining structures (RFERS) was developed. Closed‐form equations were derived such that if one value of displacement or pressure is known (or assumed) the other can be computed for hydrostatic, seismic, uniform, and semi‐elliptical earth pressure distributions. Additionally, the general form of the equations can be used to predict the magnitude of the lateral force even if the shape of the earth pressure is unknown, with a reasonable degree of accuracy. The expressions for deflection were derived by treating the structure as an equivalent cantilever beam and calibrating the resulting expression using the finite element method (FEM). A parametric FEM analysis, of 42 000 different RFERS configurations, was performed to calibrate the expressions, using multivariate non‐linear regression between the derived expressions and FEM. A Weibull statistical analysis was performed for each equation and determined that the equations had better than 80% probability to yield deflections that are within 25% of the value computed using FEM. Furthermore, there is a 98% certainty that each equation will yield a deflection that is within 50% of that computed using FEM. Copyright © 2011 John Wiley & Sons, Ltd.     

分离草虾杆状病毒包涵体的一种新方法

本文报道了一种纯化草虾杆状病毒（MBV）包涵体的方法。取出受MBV感染的虾肝胰组织，在TESP溶液中（50mmol/dm^3Tris-HC1 pH8.0,50mmol/dm^3EDTA,0.1mol/dm^3NaC1,0.5mmol/dm^3PMSF）破碎，差速离心，再经30%～70%蔗糖浓度梯度高速离心，可得到含包涵体的区带。电镜下观察到较纯的包涵体，经蛋白酶K降解后，PCR扩增，经琼脂糖凝胶电泳可见清晰的DNA区带。