The Medium-Resolution Spectral Imager-Ⅱ(MERSI-Ⅱ)instrument aboard China’s Fengyun-3D satellite shares similarities with NASA’s Moderate Resolution Imaging Spectroradiometer(MODIS)sensor,enabling the retrieval of g...The Medium-Resolution Spectral Imager-Ⅱ(MERSI-Ⅱ)instrument aboard China’s Fengyun-3D satellite shares similarities with NASA’s Moderate Resolution Imaging Spectroradiometer(MODIS)sensor,enabling the retrieval of global aerosol optical depth(AOD).However,no officially released operational MERSI-Ⅱ aerosol products currently exist over the ocean.This study focuses on adapting the MODIS dark target(DT)ocean algorithm to the MERSI-Ⅱ sensor.A retrieval test is conducted on the 2019 MERSI-Ⅱ data over the global ocean,and the retrieved AODs are validated against ground-based measurements from the automatic Aerosol Robotic Network(AERONET)and the shipborne Maritime Aerosol Network(MAN).The operational MODIS DT aerosol products are also used for comparison purposes.The results show that MERSI-Ⅱ AOD granule retrievals are in good agreement with MODIS products,boasting high correlation coefficients(R)of up to 0.96 and consistent spatial distribution trends.Furthermore,the MERSI-Ⅱ retrievals perform well in comparison to AERONET and MAN measurements,with high R-values(>0.86).However,the low-value retrievals from MERSI-Ⅱ tend to be slightly overestimated compared to MODIS,despite both AODs displaying a positive bias.Notably,the monthly gridded AODs over the high latitudes of the northern and southern hemispheres suggest that MERSI-Ⅱ exhibits greater stability in space and time,effectively reducing unrealistically high-value noise in the MODIS products.These results illustrate that the MERSI-Ⅱ retrievals meet specific accuracy requirements by maintaining the algorithmic framework and most of the algorithmic assumptions,providing a crucial data supplement for aerosol studies and climate change.展开更多
由于缺乏大规模的雾天飞机目标遥感数据集,现有的目标检测方法难以在雾天条件下实现高精度的目标识别和定位任务。针对这一问题,提出了一种雾天条件下飞机目标检测方法,该方法结合了暗通道先验算法和Faster R⁃CNN(Faster Regions with C...由于缺乏大规模的雾天飞机目标遥感数据集,现有的目标检测方法难以在雾天条件下实现高精度的目标识别和定位任务。针对这一问题,提出了一种雾天条件下飞机目标检测方法,该方法结合了暗通道先验算法和Faster R⁃CNN(Faster Regions with Convolutional Neural Network Features)模型。首先,随机选取少量飞机目标原始图像,通过图像处理数据增强法扩展原始图像遥感数据集。其次,利用暗通道先验算法计算真实雾气图像的透射率值,并将其移植到原始图像中,生成雾气模拟的遥感数据集。最后,使用创建的数据集训练Faster R⁃CNN网络模型以完成飞机目标的识别和定位任务。实验结果表明,与原始数据集相比,该数据集在轻雾和浓雾状态下的检测性能都有明显提高,证明了所提数据集对于雾天环境下飞机目标检测的有效性和实用性。展开更多
基金supported in part by the National Natural Science Foundation of China(Grant Nos.42471424,41975036,and 42075132)the Fengyun Application Pioneering Project(Grant No.FY-APP024)+1 种基金the State Key Project of National Natural Science Foundation of China-Key projects of joint fund for regional innovation and development(Grant No.U22A20566)the Scientific and Technological Innovation Team of Universities in Henan Province(Grant No.22IRTSTHN008).
文摘The Medium-Resolution Spectral Imager-Ⅱ(MERSI-Ⅱ)instrument aboard China’s Fengyun-3D satellite shares similarities with NASA’s Moderate Resolution Imaging Spectroradiometer(MODIS)sensor,enabling the retrieval of global aerosol optical depth(AOD).However,no officially released operational MERSI-Ⅱ aerosol products currently exist over the ocean.This study focuses on adapting the MODIS dark target(DT)ocean algorithm to the MERSI-Ⅱ sensor.A retrieval test is conducted on the 2019 MERSI-Ⅱ data over the global ocean,and the retrieved AODs are validated against ground-based measurements from the automatic Aerosol Robotic Network(AERONET)and the shipborne Maritime Aerosol Network(MAN).The operational MODIS DT aerosol products are also used for comparison purposes.The results show that MERSI-Ⅱ AOD granule retrievals are in good agreement with MODIS products,boasting high correlation coefficients(R)of up to 0.96 and consistent spatial distribution trends.Furthermore,the MERSI-Ⅱ retrievals perform well in comparison to AERONET and MAN measurements,with high R-values(>0.86).However,the low-value retrievals from MERSI-Ⅱ tend to be slightly overestimated compared to MODIS,despite both AODs displaying a positive bias.Notably,the monthly gridded AODs over the high latitudes of the northern and southern hemispheres suggest that MERSI-Ⅱ exhibits greater stability in space and time,effectively reducing unrealistically high-value noise in the MODIS products.These results illustrate that the MERSI-Ⅱ retrievals meet specific accuracy requirements by maintaining the algorithmic framework and most of the algorithmic assumptions,providing a crucial data supplement for aerosol studies and climate change.
文摘由于缺乏大规模的雾天飞机目标遥感数据集,现有的目标检测方法难以在雾天条件下实现高精度的目标识别和定位任务。针对这一问题,提出了一种雾天条件下飞机目标检测方法,该方法结合了暗通道先验算法和Faster R⁃CNN(Faster Regions with Convolutional Neural Network Features)模型。首先,随机选取少量飞机目标原始图像,通过图像处理数据增强法扩展原始图像遥感数据集。其次,利用暗通道先验算法计算真实雾气图像的透射率值,并将其移植到原始图像中,生成雾气模拟的遥感数据集。最后,使用创建的数据集训练Faster R⁃CNN网络模型以完成飞机目标的识别和定位任务。实验结果表明,与原始数据集相比,该数据集在轻雾和浓雾状态下的检测性能都有明显提高,证明了所提数据集对于雾天环境下飞机目标检测的有效性和实用性。
