论文部分内容阅读
合成孔径雷达(SAR)数据对于南方多云多雨天气的地表农作物类型的探测具有独特的优势。以江苏省海安县为例,基于多极化SAR数据,包括双极化ALOS PALSAR以及全极化Radarsat-2数据,采用面向对象的方法,针对当地水稻/旱田进行识别。针对双极化SAR数据,利用了其强度信息进行分类识别;而基于全极化数据,除强度信息外,还利用了其SAR信号统计分布概率进行分类规则建立。结果表明:L波段的ALOS PALSAR在识别旱地的桑树方面具有很大的优势,而基于两种分类方法的C波段Radarsat-2数据识别水稻的精度分别为85%和75%,略低于ALOSPALSAR的识别结果(87.5%)。
Synthetic aperture radar (SAR) data have unique advantages for the detection of surface crop types in the southern, cloudy, and rainy weather. Taking Hai’an County, Jiangsu Province as an example, an object-oriented method was used to identify local rice / upland fields based on multi-polarization SAR data, including dual-polarized ALOS PALSAR and fully-polarized Radarsat-2 data. For the dual-polarized SAR data, the intensity information is used for classification and identification. Based on the data of the total polarization, in addition to the intensity information, the statistical distribution probability of SAR signals is used to establish classification rules. The results showed that ALOS PALSAR in L band had a great advantage in identifying mulberry trees in dryland, while the accuracy of C-band Radarsat-2 data identification in rice based on the two classification methods was 85% and 75%, respectively, slightly lower than those in ALOSPALSAR Identification results (87.5%).