实时GIS时空数据模型

为满足动态目标与传感器等实时观测数据获取、存储、管理、分析与可视化的要求,需要发展一种新型地理信息系统—实时GIS。本文根据实时GIS中各种地理要素的特点以及存贮管理要求,提出了一种面向动态地理对象与动态过程模拟的实时GIS时空数据模型,它将时空过程、地理对象、事件、事件类型、状态、观测等相关要素整合在一个时空数据模型中。基于该模型研发了新一代实时GIS,并以四种动态地理对象（包括移动对象、原位传感器对象、视频对象和过程模拟对象）的时空数据的实时接入、存储与可视化为例,验证的模型的可行性。     

High Resolution Still Video Camera for Industrial Photogrammetry

Still video cameras are easy to handle, portable, digital image acquisition devices for close range photogrammetry. They allow fast object recording without being connected to a computer. As an example of the new camera type, the high resolution Kodak DCS200 is presented in this paper, combined with the DPA digital photogrammetric station used for image measurement and object reconstruction. The performance of this three dimensional measurement system is illustrated by calibration results as well as applications to high precision dimensional inspection in industry.     

监控视频中动态目标与地理空间信息的融合与可视化方法

监控视频的动态前景目标智能分析是平安城市、智慧园区等安防建设的重要基础,将监控视频与地理空间数据融合可为静态的地理数据赋予动态属性。针对传统监控视频与地理信息数据集成仅仅将视频数据投射至地理空间,造成存储难、视频内容理解难度大等问题,本文提出了前景动态目标与地理空间信息的融合模型,通过推导出的映射模型将图像空间中的动态前景目标及跟踪轨迹映射至地理空间中,达到将监控视频与地理信息有机融合的目的。根据不同的应用需求,本文设计了4种多图层融合显示模式,实现了监控视频中的动态前景目标在地理空间的可视化。     

一种能适应月相变化的月球图像中心提取算法

月球是夜间最亮的天体,测月导航可在云雾天气下有效弥补传统天文导航在夜间只能观测恒星的不足,扩大天文导航的可用范围。为实现测月导航,提出了一种能适应不同月相变化的月球敏感器图像月球中心提取算法。利用Sobel-Zernike矩算子对图像进行边缘检测,利用基于圆形性检测的搜索算法筛选月球边缘点并以此提取月球中心。经半物理仿真实验解算,该算法能够针对不同月相图像有效提取月球中心,中心提取的内符合平均精度为0.16像素,外符合平均精度为2.86″。由于在月球上观测地球也有"地相"变化,该算法同样适用于月球车地球敏感器图像中心提取。     

Sensor modeling,self-calibration and accuracy testing of panoramic cameras and laser scanners

Terrestrial Linear Array CCD-based panoramic cameras have been used for purely imaging purposes, but they also have a high potential for use in high accuracy measurement applications. The imaging geometry and the high information content of those images make them suitable candidates for quantitative image analysis. For that a particular sensor model has to be established and the inherent accuracy potential has to be investigated. We developed a sensor model for terrestrial Linear Array-based panoramic cameras by means of a modified bundle adjustment with additional parameters, which models substantial deviations of a real camera from the ideal one. We used 3D straight-line information in addition to tie points to conduct a full calibration and orientation without control point information. Due to the similarity of the operation of laser scanners to panoramic cameras the sensor model of the panoramic cameras was extended for the self-calibration of laser scanners. We present the joint sensor model for panoramic cameras and laser scanners and the results of self-calibration, which indicate a subpixel accuracy level for such highly dynamic systems. Finally we demonstrate the systems’ accuracy of two typical panoramic cameras in 3D point positioning, using both a minimal number of control points and a free network adjustment. With these new panoramic imaging devices we have additional powerful sensors for image recording and efficient 3D object modeling.     

一种激光厚度测量方法的研究

阐述了利用激光器、图像传感器进行非接触厚度自动测量的原理。实验表明该方法具有实时性能好，测量重复精度较高等优点     

DIGITAL IMAGE CORRELATION: PERFORMANCE AND POTENTIAL APPLICATION IN PHOTOGRAMMETRY

A procedure for digital image correlation is described which is based on least squares window matching. The immediate aim is high precision parallax assessment, point transfer, and point measurement. Experiments and theory have confirmed the high accuracy potential of the method. By implementation of charge coupled device (CCD) video cameras in an analytical plotter, an experimental hardware and software configuration has been established with which the operational on line application of digital image correlation for conventional photogrammetric measuring tasks can be tested. First results of calibration and performance of the system are presented. They allow optimistic conclusions as to the further development and practical application of digital image processing in photogrammetry.     

运动视频序列影像的自动配准

影像配准是运动视频处理应用中的关键技术,从计算机视觉角度出发提出了一种针对运动视频处理的自动影像配准方法。其主要思路可以归结为基于特征匹配的配准过程,具体分为3个方面:利用Harris算子检测角点特征;以互相关函数为测度对角点特征进行初步匹配,特别使用RANSAC拟合基础矩阵F的方法剔除匹配中的错误对应;利用得到的结果重新拟合配准模型进行重采样变换。最后进一步分析了此方法应用于具体运动视频处理时需要考虑的一些问题。     

A LARGE SCALE DEFORMATION STUDY USING MOIRÉ PHOTOGRAPHY

This article describes what is probably the largest application of a moiré technique. Real time video systems and a remotely controlled, high quality still camera were used to monitor the buckling in the walls of a liquid gas storage tank, in a section more than 30 m across.     

一种基于道路监控视频图像质量评估的新方法

针对近些年图像质量评价的重要性,提出一种针对道路监控视频的图像质量评估方法。因为道路监控视频具有连贯性、图像稳定性强、单个监控视频场景单一等特点,本文采用一种结合多幅图像创建物理模型对比恢复天气影响的图像质量评估方法,通过多幅图像比较恢复道路视频图像清晰度,加入噪声亮度比的因子来增加图像的鲁棒性和准确性,分析该方法的优势和不足。     

顾及像面畸变的卫星视频稳像

卫星视频每帧影像外方位元素数值存在差别,相邻帧影像间的同名点在像面位置不同,且由于每帧影像均存在几何畸变,经典的帧间的投影变换模型、仿射变换模型等视频稳像的配准模型不能精确描述帧间的同名点关系。针对该问题,本文提出顾及像面畸变的卫星视频稳像算法,基于仿真的卫星视频,验证了该算法的可行性和精度。     

面向关系数据库扩展的自适应影像金字塔模型

随着遥感技术的发展,特别是传感器分辨率的提高,影像数据量日益增长,这对高效、智能地管理海量数据提出了更高的要求,如何有效、智能地存取海量遥感数据成为当前越来越热门的话题。本文针对空间数据管理系统的发展现状,结合数据库存储、网络传输和客户端的硬件情况,从影像数据切片和影像数据的细节层次(level of detail,LOD)模型两个层面来解决遥感数据的存储问题,提出了自适应的遥感影像切片机制,以及与之相适应的LOD模型构建方法。通过试验证实,该影像管理机制能够在多种不同的数据库、网络和客户端条件下实现影像数据智能化高效存储与访问。本文的研究为高效的影像数据管理提供了可行的思路和方法,有助于在C/S架构的网络环境下实现基于数据库技术的遥感影像数据的高效存取和管理。     

数码相机实施摄像测量的几个问题

提出像方几何量均取像素为单位处理，以直接使用现有摄影测量程序并使检校过程简化，给出α值的检校要求关系式，罗列主距的锁定方式，总结数码相机所摄影像的多种处理方法并充分利用数码相机功能。     

The Metric Impact of Reduction Optics in Digital Cameras

The recently released Nikon E2 still video camera incorporates reduction optics technology which enables a full conventional 35mm image to be recorded on a 2/3 inch (17mm) CCD array. This paper investigates the metric impact of reduction optics through application of the technique of multisensor system self-calibration. The influence of the JPEG image compression facility contained within the E2 is also assessed in an overall evaluation of the camera for vision metrology applications which demand relatively high object point triangulation accuracies.     

An improved Bag-of-Words framework for remote sensing image retrieval in large-scale image databases

Due to advances in satellite and sensor technology, the number and size of Remote Sensing (RS) images continue to grow at a rapid pace. The continuous stream of sensor data from satellites poses major challenges for the retrieval of relevant information from those satellite datastreams. The Bag-of-Words (BoW) framework is a leading image search approach and has been successfully applied in a broad range of computer vision problems and hence has received much attention from the RS community. However, the recognition performance of a typical BoW framework becomes very poor when the framework is applied to application scenarios where the appearance and texture of images are very similar. In this paper, we propose a simple method to improve recognition performance of a typical BoW framework by representing images with local features extracted from base images. In addition, we propose a similarity measure for RS images by counting the number of same words assigned to images. We compare the performance of these methods with a typical BoW framework. Our experiments show that the proposed method has better recognition performance than that of the BoW and requires less storage space for saving local invariant features.     

机载LiDAR数据和航空影像基于角特征的配准

机载激光雷达(Light Detection And Ranging,LiDAR)系统近年来逐渐成为一种新兴的测量手段,是一种可以直接提供大量地面离散点的高精度三维信息的主动式遥感技术,具有依赖控制测量条件少、受天气影响小、自动化水平高等优点。然而LiDAR获取的数据主要是离散激光点云,具有高精度的空间结构信息而缺少纹理和光谱信息。将LiDAR数据和遥感影像数据进行融合将是当今测绘科学研究的方向之一。融合影像数据和点云数据的关键步骤就是将两者进行高精度的配准。     

多镜头组合型全景相机两种成像模型的定位精度比较

根据多镜头组合型全景相机的严格成像机理,分析了广泛采用的全景理想成像模型系统误差的来源,推导出误差分布规律,并深入比较了严格成像模型和理想成像模型之间的差异.首先,分析了理想模型的单片物方定位误差和像方后方交会误差,理论公式表明其误差分布和物像间距及视角大小相关;其次,分析了全景立体像对的理想核线误差,并证明误差受像对间的旋转角和Z坐标影响.最后,分别采用室内、外数据对两种模型的空间后方交会、核线模型、光束法平差进行比较分析.试验结果表明,室外条件下由于物距接近于全景投影面、立体影像间旋转角较小等因素,两种模型的差异很小,且都达到了相当于1像素左右的测量精度;在狭窄的室内条件下,由于物距小、旋转角大等因素,严格模型在测量精度上明显优于理想模型.     

敦煌场地CBERS-02 CCD传感器在轨绝对辐射定标研究

对资源一号卫星02星（CBERS-02）,为深入其定量化分析,定期在敦煌绝对辐射校正场地进行了地面同步测量,开展了更新在轨绝对辐射定标研究。另外,运用地表反射率和大气光学特性参量数据,计算出CBERS-02CCD传感器4个波段的绝对辐射标定系数。同时,利用2004年8月25目的Landsat-5 TM图像数据对CBERS-02 CCD传感器的辐射特性进行交叉定标,验证分析了CBERS-02 CCD传感器场地绝对辐射定标的基本特性及其可靠性。另外,为CBERS后续星的传感器改进,提供了分析参量。     

小波低频子带预测的遥感图像分布式编码

传感技术的不断发展使遥感图像的时间、空间和光谱分辨率不断提高,信息量日益增大,给遥感图像的存储和传输带来了极大挑战,传统图像编码策略难以适应这种基于上行的遥感图像应用需求。本文提出一种基于小波低频子带预测的分布式遥感图像编码方案,利用预测插值的方法得到用于解码Wyner-Ziv的辅助信息,图像的低频信息采用小波零树编码方案,而预测误差图像采用基于均匀量化和位平面的Turbo混合编码方案;在解码端,以预测误差图像的码流作为辅助信息,将其与图像的低频信息码流进行联合解码。实验结果表明,提出方法具有编码效率高、失真率小、压缩比高、抗误码性强和码流具有质量可分级等特性。     

交通监控视频图像语义分割及其拼接方法

视频拼接是图像拼接的外延,在场景监控、目标识别等应用中发挥着重要作用。传统视频拼接算法多要求视频间具有较大重叠区域且特征点匹配过程中只顾及图像几何特征,当处理交通监控视频时,会因不同摄像头之间重叠区域极小或主光轴之间夹角较大而导致无法拼接或图像变形较大。为此,本文提出一种交通监控视频图像语义分割及其拼接方法。首先,利用边缘角度二阶差分直方图算法自动识别多视频交汇区域的正射影像,并将其作为拼接背景图像;然后,基于全卷积神经网络对正射影像和视频图像进行语义分割,提取图像中的交通专题语义;最后,以交通专题语义作为约束进行特征点匹配,将各个交通监控视频匹配至背景正射影像,实现监控区域视频拼接。采用山东省某市实际视频数据进行试验验证,结果表明对于重叠区域较小的监控视频,本文方法可获得较好地拼接图像,同时可有效提高特征点匹配的准确度。