基于CBERS-02B卫星影像光谱信息的水稻种植面积提取方法——以南京市溧水县为例

Extraction of rice planting area based on CBERS-02B image spectral information:a case study in Lishui,Nanjing

利用水稻成熟期的影像,在进行光谱信息的统计和分析的基础上,对比水稻与其他植被在各波段DN(digital number)值以及NDVI(normalized difference vegetable index)值的数值差异,总结出水稻与其他植被反射光谱特征,认为成熟期的水稻在红光、绿光、蓝光波段DN值以及NDVI值上具有异于其他植被的特征,并通过波段运算扩大其数值差异,建立水稻像元提取模型,确定阈值,从而进行水稻种植面积的提取,精度可达91.95%。该方法能满足小区域的水稻种植面积提取和制图,并能够满足一定的精度要求。

The CBERS-02B image at rice maturity stage was used in this study.Based on the statistics and analyses of the image spectral information,the general regularity and characters of rice spectral reflectance were summarized.It indicated that,at the rice maturity stage,the characters of rice were different between the others′ in the values of red,green,blue bands and NDVI.Bands calculation was used to build the relationships,and the thresholds were defined.Finally the rice planting area was extracted by the relationships and the thresholds.The accuracy reached 91.95%.This method can improve the precision of rice planting area interpretation of remote sensing images.