首页> 外文会议>Image and Signal Processing for Remote Sensing III >Some applications of SAR image registration: autofocusing odometry moving-object localization and navigation resetting
【24h】

Some applications of SAR image registration: autofocusing odometry moving-object localization and navigation resetting

机译:SAR图像配准的一些应用:自动聚焦测距法运动对象定位和导航重置

获取原文

摘要

Abstract: Though accurate motion compensation for SAR satelliteimaging is not an acute problem at present, it willprobably become one in the near future with increase inthe ground resolution of available sensors. Thisproblem is already nowadays extremely pertinent forairborne SAR imaging, especially with slow and/orunmanned aircrafts. Image registration is often thefirst and highly critical step in many operational usesof SAR imaging. We illustrate some of thoseapplications with highly distorted airborne SAR images.Those applications are of two main classes, one beingthe registration of two similar SAR images without anexternal reference, and the second being theregistration of a SAR image with another type ofgeographical data such as an optical image or map. Thelocal similarity of the images in the first class ofproblems leads us to adapt a technique from opticalflow motion evaluation. We used a local correlationtechnique with a low dimensional parametrization of themotion to be estimated. The second class of problems isdealt by first extracting linear (sometimes pointwiseor areal) features and then matching them by means of amulti hypothesis filtering. We shall describe here fourapplications of image registration: (1) The first oneis the registration of single-look images obtainedwithin the same overflight in order to compensate fortrajectory errors not detected by the aircraftnavigation system. This registration not only allowsthe multi-look incoherent summation, but also therefocusing of the single-look images to increaseazimuth resolution. This process is the type of'auto-focusing' known as frame-drift. (2) Multi-lookregistration also provides an accurate evaluation ofground speed of the carrier aircraft that can beexploited for internal navigation unit hybridization,even in the absence of cartographic knowledge of thearea flown over. This application is called'odometering.' (3) Multi-look registration isfurthermore extremely critical for moving vehiclesdetection and their proper motion estimation usingmulti-look imaging on wide beam antennas. This is oneversion of the problem known as 'moving targetindication.' (4) Registration of a SAR image with a mapor an optical image of a given area is also used foraccurate localization of objects detected ingeographical coordinates, for their designation toother systems. It may also be used for accuratelylocalizing the carrier air-vehicle, especially whenunmanned. This is the 'navigation resetting problem.'!9
机译:摘要:虽然目前对SAR卫星成像进行精确的运动补偿并不是一个迫切的问题,但是随着可用传感器的地面分辨率的提高,它可能会在不久的将来成为一种。如今,这个问题已经与机载SAR成像极为相关,尤其是在慢速和/或无人驾驶飞机上。图像配准通常是SAR成像的许多操作用途中的第一步,也是至关重要的一步。我们用高度失真的机载SAR图像说明了其中一些应用,这些应用主要分为两类,一类是没有外部参考的两个相似SAR图像的配准,第二类是具有另一种地理数据(例如光学)的SAR图像的配准。图片或地图。第一类问题中图像的局部相似性导致我们采用光流运动评估中的一种技术。我们使用具有低维运动参数化的局部相关技术来进行估计。第二类问题是处理问题,方法是先提取线性(有时是点向或面)特征,然后通过多假设过滤将其匹配。我们将在这里描述图像配准的四个应用:(1)第一个是在同一飞行中获得的单视图像的配准,以补偿飞机导航系统未检测到的轨迹误差。该配准不仅允许多视点不相干求和,而且还可以使单视点图像聚焦以增加方位角分辨率。此过程是称为帧漂移的“自动聚焦”类型。 (2)多方位配准还可以准确评估舰载飞机的地面速度,即使在没有流经区域的制图知识的情况下,也可以将其用于内部导航单元混合。该应用称为“里程表”。 (3)此外,多视点配准对于移动车辆的检测以及使用宽波束天线上的多视点成像进行正确的运动估计至关重要。这是称为“移动目标指示”的问题的一个版本。 (4)将SAR图像与测绘仪配准,将给定区域的光学图像用于准确定位检测到的地理坐标的对象,并将其指定给其他系统。它也可用于准确定位运载工具的车辆,尤其是在无人驾驶时。这是“导航重置问题”!9

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号