目的全光场相机一次曝光获得4维光场信息,即2维空间信息和2维角度信息。然而,渲染获取图像的空间分辨率较低,角度分辨率也无法满足应用需求。针对此问题,提出一种边缘结构保持加权的超数字化全变分(BDTV)超分辨率重建算法。方法将重建...目的全光场相机一次曝光获得4维光场信息,即2维空间信息和2维角度信息。然而,渲染获取图像的空间分辨率较低,角度分辨率也无法满足应用需求。针对此问题,提出一种边缘结构保持加权的超数字化全变分(BDTV)超分辨率重建算法。方法将重建后图像与不同视角低分辨率图像的边缘拟合误差函数作为数据项的权重,并利用BDTV先验模型去噪和保持边缘。结果为验证本文方法的有效性,将本文方法与其他方法进行了实验,其结果是无论是在视觉效果还是指标评价方面,本文方法的重建效果和边缘保持效果都优于其他方法,峰值信噪比提升约1 d B,结构相似性指数也有约0.01的提升。结论本文算法不仅提高了全光场图像的空间分辨率和角度分辨率,同时也更好地保持了图像的边缘信息,特别是当深度图存在误差时,该算法的超分辨率重建和边缘保持效果更为明显。展开更多
FY-4 is the second generation of Chinese geostationary satellite for quantitative remote sensing meteorological application. The detection efficiency, spectral bands, spatial and time resolution have been greatly impr...FY-4 is the second generation of Chinese geostationary satellite for quantitative remote sensing meteorological application. The detection efficiency, spectral bands, spatial and time resolution have been greatly improved with respect to those of first generation, as well as the radiometric calibration and sensitivity. The combination of multichannel detection and vertical sounding was first realized on FY-4, because both the Advanced Geostationary Radiation Imager(AGRI) and Geostationary Interferometric Infrared Sounder(GIIRS) are on the same spacecraft. The main performance of the payloads including AGRI, GIIRS and Lightning Mapping Imager, and the spacecraft bus are presented, the performance being equivalent to the level of the third generation meteorological satellites in Europe and USA. The acquiring methods of remote sensing data including multichannel and high precision quantitative observing, imaging collection of the ground and cloud, vertical observation of atmospheric temperature and moisture, lightning imaging observation and space environment detection are shown. Several innovative technologies including high accuracy rotation angle detection and scanning control, high precision calibration, micro vibration suppression, unified reference of platform and payload and on-orbit measurement, real-time image navigation and registration on-orbit were applied in FY-4.展开更多
文摘目的全光场相机一次曝光获得4维光场信息,即2维空间信息和2维角度信息。然而,渲染获取图像的空间分辨率较低,角度分辨率也无法满足应用需求。针对此问题,提出一种边缘结构保持加权的超数字化全变分(BDTV)超分辨率重建算法。方法将重建后图像与不同视角低分辨率图像的边缘拟合误差函数作为数据项的权重,并利用BDTV先验模型去噪和保持边缘。结果为验证本文方法的有效性,将本文方法与其他方法进行了实验,其结果是无论是在视觉效果还是指标评价方面,本文方法的重建效果和边缘保持效果都优于其他方法,峰值信噪比提升约1 d B,结构相似性指数也有约0.01的提升。结论本文算法不仅提高了全光场图像的空间分辨率和角度分辨率,同时也更好地保持了图像的边缘信息,特别是当深度图存在误差时,该算法的超分辨率重建和边缘保持效果更为明显。
文摘FY-4 is the second generation of Chinese geostationary satellite for quantitative remote sensing meteorological application. The detection efficiency, spectral bands, spatial and time resolution have been greatly improved with respect to those of first generation, as well as the radiometric calibration and sensitivity. The combination of multichannel detection and vertical sounding was first realized on FY-4, because both the Advanced Geostationary Radiation Imager(AGRI) and Geostationary Interferometric Infrared Sounder(GIIRS) are on the same spacecraft. The main performance of the payloads including AGRI, GIIRS and Lightning Mapping Imager, and the spacecraft bus are presented, the performance being equivalent to the level of the third generation meteorological satellites in Europe and USA. The acquiring methods of remote sensing data including multichannel and high precision quantitative observing, imaging collection of the ground and cloud, vertical observation of atmospheric temperature and moisture, lightning imaging observation and space environment detection are shown. Several innovative technologies including high accuracy rotation angle detection and scanning control, high precision calibration, micro vibration suppression, unified reference of platform and payload and on-orbit measurement, real-time image navigation and registration on-orbit were applied in FY-4.
