无人机原始影像快速拼接系统的设计与实现  

Design and implementation of UAV original images quick mosaicing system

作　　者：刘如飞[1] 刘冰[1] 卢秀山[1] 田茂义[1] 

机构地区：[1]山东科技大学测绘科学与工程学院,山东青岛266590

出　　处：《测绘科学》2013年第5期170-171,182,共3页Science of Surveying and Mapping

基　　金：2011年中国冶金地质总局科技研发项目(zykj201102);国家科技支撑计划项目(2011BAK07B01)

摘　　要：目前无人机原始影像拼接效率不高,影响了应急指挥和应用。针对此问题,本文通过分析影像特点,提出应用改进SIFT算法和GIS组件技术实现拼接系统的方法:为消除无控制点原始影像拼接的误差累积问题,提出双向与分区拼接,旋转与缩放纠正等拼接策略,并利用GIS影像管理方法实现影像坐标连接和综合应用。实验表明,拼接方法既保证了拼接的精度,又提高了拼接的效率和实际应用价值。At present, the mosaic of UAV original image is inefficient, which affects the emergency command and application. The improved SIFT algorithm and GIS component technology were proposed to realize the mosaic system through the analysis of image characteristics. The mosaic methods of bidirectional and subregion mosaicing, rotation and scaling correction were proposed to eliminate the accumulated errors of original image mosaicing without GCP （ Ground Control Point）. Moreove GIS image management method was used to achieve the image coordinate connection and comprehensive application. The experimental results showed that the mosaic method could both ensure the accuracy of mosaic, and improve the efficiency of mosaic and the practical applying value.

关 键 词：SIFT GIS组件 影像拼接策略 

分 类 号：P208[天文地球—地图制图学与地理信息工程] P237[天文地球—测绘科学与技术]