论文部分内容阅读
无人飞行器低空遥感监测影像的拼接应具备快速、高效等特点,D.G.Lowe提出的基于SIFT特征匹配全景影像自动镶嵌算法无需人工交互,具有很强的鲁棒性,但是全景镶嵌算法主要是针对传感器在固定位置围绕竖直轴向360°旋转采集的序列影像,因此,并不适应用于无人飞行器遥感监测影像的拼接。根据无人飞行器遥感监测在弱透视投影条件下的成像模型和影像数据是按照设定航线等间距获取等特点,对基于SIFT特征匹配的影像自动镶嵌算法在影像空间变换等方面进行改进,同时增加了畸变校正、边缘裁切等预处理方法,形成完整的无人飞行遥感监测影像拼接流程,并通过试验证明该算法的可行性。
UAV low-altitude remote sensing image mosaic should be fast and efficient features, DGLowe proposed SIFT feature matching panoramic image automatic mosaic algorithm without human interaction, with strong robustness, but the panoramic mosaic algorithm is mainly for sensors The sequence of images acquired at 360 ° rotation around a vertical axis in a fixed position is therefore not suitable for splicing for remotely piloted surveillance images of UAVs. According to the characteristics of the imaging model and image data of unmanned aerial vehicle remote sensing in the perspective of weak projection, the automatic image mosaic algorithm based on the SIFT feature matching is improved in terms of image space transformation and the like, Distortion correction, edge cutting and other pretreatment methods to form a complete unmanned aerial remote sensing monitoring image mosaic process, and the feasibility of the algorithm is proved through experiments.