亚热带资源与环境学报
亞熱帶資源與環境學報
아열대자원여배경학보
JOURNAL OF SUBTROPICAL RESOURCES AND ENVIRONMENT
2014年
2期
82-87
,共6页
茶树%光谱%端元%组分%像元
茶樹%光譜%耑元%組分%像元
다수%광보%단원%조분%상원
Camellia sinensis%spectrum%endmember%composition%pixel
准确提取端元并确定像元中端元组分和丰度是提高遥感分类精度的基础和前提。本研究根据福建省多数茶园特点,从对地观测角度将茶园分为垄前冠面、垄后冠面、沟中阴影面和沟中光照面4种端元,并测定各类端元光谱特征。根据光线入射方向和地形参数计算茶园中任意端元组合面的丰度,利用几何光学模型进行组分光谱混合模拟,最后把模拟结果与实测结果相比较。研究结果发现,模拟和实测的光谱曲线很接近,光谱的反射峰、吸收谷位置基本吻合,总体精度在85%以上,说明采用组分合成法模拟茶园光谱是可行的。
準確提取耑元併確定像元中耑元組分和豐度是提高遙感分類精度的基礎和前提。本研究根據福建省多數茶園特點,從對地觀測角度將茶園分為壟前冠麵、壟後冠麵、溝中陰影麵和溝中光照麵4種耑元,併測定各類耑元光譜特徵。根據光線入射方嚮和地形參數計算茶園中任意耑元組閤麵的豐度,利用幾何光學模型進行組分光譜混閤模擬,最後把模擬結果與實測結果相比較。研究結果髮現,模擬和實測的光譜麯線很接近,光譜的反射峰、吸收穀位置基本吻閤,總體精度在85%以上,說明採用組分閤成法模擬茶園光譜是可行的。
준학제취단원병학정상원중단원조분화봉도시제고요감분류정도적기출화전제。본연구근거복건성다수다완특점,종대지관측각도장다완분위롱전관면、롱후관면、구중음영면화구중광조면4충단원,병측정각류단원광보특정。근거광선입사방향화지형삼수계산다완중임의단원조합면적봉도,이용궤하광학모형진행조분광보혼합모의,최후파모의결과여실측결과상비교。연구결과발현,모의화실측적광보곡선흔접근,광보적반사봉、흡수곡위치기본문합,총체정도재85%이상,설명채용조분합성법모의다완광보시가행적。
Accurate extraction of endmembers and calculation of their composition and abundance in each pixel are the foundation as well as prerequisite for increasing classification accuracy of remote sensing .According to the general features of tea plantations in Fujian , endmembers of tea plantations were divided into four kinds , including surface of front canopy , surface of behind canopy , shadow area of ditch and illumination area of ditch .Then, the spectral reflectance of each endmember was meas-ured, and the abundance of each combination of tea plantations was calculated by incident light direc -tions and topographic parameters .Spectra of tea plantations of mixed endmembers were simulated using GO model.And, the simulation spectrum was compared with that in situ .By comparing the results , it was found that the two kinds of spectral curve were very similar to each other and the reflection peak or valley locations were close .The total simulation accuracy of the model was above 85%, showing the feasibility of GO model in tea plantations .