摘要
遥感影像几何校正的质量和速度将会直接影响到后续的数据处理和定量提取信息的质量。基于大数据量的3级无人机影像,探讨将其从西安80坐标转换到北京54坐标的快速处理方法。分别采用Erdas二次多项式变换和FME仿射变换进行单幅影像校正,对比发现仿射变换校正后的影像与参考卫星影像地物匹配效果更好,可达到应用的精度。并利用FME批处理功能,大大提高了工作效率。实验表明:当控制点精度满足要求且原始到目标坐标系的转换属于二维线性变换时,FME是快速几何校正的一种有效手段。
The quality and speed of geometric correcting remote sensing image directly affect the subsequent data processing and the quality of quantitative information extraction. Based on large amounts of 3-level UAV images, the rapid conversion approach from the Xi＇ an 80 coordinate to Beijing 54 coordinate is discussed on. Single image correction is done by using quadratic polynomial transformation of Erdas and affine transformation of FME. It is found that the feature matching of corrected image using affine transformation is better with the reference satellite image. The application of precision is achieved. And using FME batch function, the work efficiency is greatly improved. The experiments show： when the control point accuracy is meeting the requirements and the coordinate system conversion from the original to the target is two-dimensional linear transform, FME is an effective mean of rapid geometric correction.
出处
《信息技术》
2014年第2期23-26,30,共5页
Information Technology
基金
国土资源部公益性行业科研专项经费项目(201111013-06)
关键词
无人机影像
几何校正
FME
仿射变换
批处理
UAV image
geometric correction
FME
affine transformation
batch processing
