论文部分内容阅读
大区域红树林多光谱遥感提取易受多种地物及自身覆盖度等因素影响,针对这一难题,文中以广西北部湾为研究区,基于HJ星CCD影像,并结合红树林空间分布特征和图谱特征,建立一种适合大区域海岸带的红树林识别提取模式.首先分析影像地物图谱特征,提出用于瞬时海水边线提取的海水比值指数,以及指示红树林的特征指数:红波段角度植被指数和像元波段反射率标准差指数.再从红树林空间生境判定和不同覆盖度反射率变化两方面入手,划分生长适宜区,通过特征指数辅助判定建立最终提取模式.应用结果表明该模式能完整准确地提取不同覆盖度红树林,最大程度消除其他干扰因素影响,总体制图精度达83.70%,用户精度为79.37%,同时也证实HJ卫星CCD多光谱遥感数据应用于海岸带红树林研究的有效性.
In view of this problem, taking Beibu Gulf of Guangxi as the research area, based on the HJ star CCD image and the spatial distribution characteristics of mangroves, the multi-spectral remote sensing extraction of mangroves in large area is susceptible to many kinds of ground objects and their own coverage. And establish a suitable mangrove extraction model suitable for the large coastal zone.Firstly, the characteristics of the atlas were analyzed, and the index of seawater ratio for the extraction of instantaneous seawater and the characteristic index of the mangroves were proposed: the red angle vegetation Index and pixel band reflectance standard deviation index.Then from the mangrove spatial habitat determination and different coverage reflectance changes two aspects, the division of growth suitable area, through the characteristic index adjuvant decision to establish the final extraction mode.Application results show that the model can The mangroves with different coverage were completely and accurately extracted to eliminate the influence of other disturbance factors. The overall mapping accuracy was 83.70% and the user accuracy was 79.37%. It was also confirmed that the HJ satellite CCD multi-spectral remote sensing data was applied to the mangrove research in coastal zone Sex.