摘要
ALOS PALSAR已经在轨运行1年多,由于其L-波段多极化观测能力,特别适合森林资源调查及生态环境相关遥感应用。然而,如何进行PALSAR数据的地理编码处理是实际应用中需要解决的一个关键问题。本文基于距离-多普勒(RD)模型发展了PALSAR Level-1.1影像产品的地球椭球地理编码校正(GEC)方法,重点论述了从PALSAR CEOS格式元数据记录中提取距离向方程、RD方程参数的具体方法。以一景已经经过几何精校正的Landsat ETM+影像为参考,获取了10个控制点,定量评价了本文所发展的GEC方法的有效性。评价结果表明:东西向均方根误差为32.529m,南北向均方根误差为32.527m,误差约是ETM+影像像元大小(28.5m)的1.14倍。本文针对PALSAR Level-1.1提出的GEC方法是进一步发展PALSAR其他数据产品的GEC和地形校正地理编码(GTC)方法的基础,对于PALSAR数据的定量化应用及其地理编码处理软件的开发具有重要的参考价值。
ALOS PALSAR has been operating in orbit for more than 1 year, its capability to observe the earth with multi-polarization mode in L-band SAR makes it much more useful for forest resource inventory and bio-environment related remote sensing applications. However, how to geocode PALSAR data is a key problem that should be solved before any practical application. A method of geocoding ellipsoid correction (GEC) for PALSAR Level-1.1 product based on Range-Doppler (RD) geolocation model was developed, and the means to derive each model parameter from PALSAR Level-1.1 product's CEOS format files was introduced in detail. Taking one scene of finely geocoded Landsat ETM+ image as reference, 10 ground control points (GCPs) were collected manually for verifying the performance of the GEC method. The root mean square error (RMSE) in West-East direction is 32. 529m, the RMSE in South-North direction is 35. 527m, both of them are about 1. 14 times of the pixel size of the reference ETM+ image. The GEC method established in this paper for PALSAR Level-1.1 product is the basis for developing the GEC and GTC (Geocoding of Terrain Correction) data processing methods for other PALSAR data at different processing levels, and is also very important and useful for the quantitative application of PALSAR and the development of relevant geocoding software.
出处
《遥感信息》
CSCD
2008年第1期37-42,95,共7页
Remote Sensing Information
基金
863项目(2006AA12010)“国家粮食主产区粮食作物种植面积遥感测量与估产业务系统”资助