中国农业机械化科学研究院集团有限公司 主管

北京卓众出版有限公司 主办

面向耕地作物种植监测的高光谱星地同步采集设计与实现

Design and implementation of hyperspectral satellite-ground synchronous acquisition for crop cultivation monitoring in cultivated land

  • 摘要: 耕地是粮食生产的根基,耕地“非农化”“非粮化”等问题屡禁不止,必须加强监测监管。高光谱遥感利用其丰富的波谱信息,拓展了耕地作物种植监测的手段。针对高光谱遥感在农业生产中耕地作物种植精细化识别的应用,存在空间分辨率不匹配、大气干扰、反演精度不稳定等问题,研究设计一套结合地面光谱仪的星地同步光谱采集技术方案,包括同步采集设计、数据处理流程、星地数据协同分析等,并以资源一号02E卫星高光谱数据为基础进行验证测试。试验结果表明,星地同步光谱采集方案的光谱一致性较高,较好地解决观测数据的尺度效应,以光谱采集为基础,农作物分类平均精度73.4%,能够较好地分析种植信息,为耕地作物生长的精细化定量反演提供数据参考。

     

    Abstract: Cultivated land is grain production's foundation.Problems such as non-agriculture and non-grain production of cultivated land have persisted despite repeated prohibitions, necessitating strengthened monitoring and supervision.Hyperspectral remote sensing, with its rich spectral information, has expanded means for crop cultivation monitoring in cultivated land.Addressing challenges in spatial resolution mismatch, atmospheric interference, and unstable inversion accuracy for hyperspectral remote sensing application in fine-scale identification of cultivated land cultivation in agricultural production, a set of satellite-ground synchronous spectral acquisition technology solutions combined with ground spectrometers was researched and designed, including synchronous acquisition design, data processing flow, and satellite-ground data collaborative analysis.Verification testing was based on hyperspectral data from ZY-1 02E satellite.Experimental results showed that satellite-ground synchronous spectral acquisition scheme achieved high spectral consistency, and effectively mitigated scale effects in observation data.Based on spectral acquisition, average accuracy of crop classification reached 73.4%, enabling effective analysis of planting information and providing data references for fine-scale quantitative inversion of crop growth in arable land.

     

/

返回文章
返回