on September 7, 2021, GaoFen5-02 (GF5-02) Satellite, of the new generation of Chinese hyperspectral remote sensing satellite was successfully launched. GF5-02, the successor to the GF5 satellite, was equipped with six advanced hyperspectral payloads. One of the most important payloads onboard the GF5-02 satellite, the Advanced Hyperspectral Imager (AHSI) has a spatial resolution of 30 m, 330 bands in a spectral range of 380-2500 nm. The spectral resolution of the visible-near infrared (VNIR) and shortwave infrared (SWIR) bands are better than 5 nm and 10 nm, respectively. In order to analyze the spectral performance of the GF5-02 AHSI, an on-orbit spectral calibration method that utilizes atmospheric limb observations with on-board calibration system was proposed in this paper. The on-orbit spectral calibration results were validated by atmospheric absorption features with synchronous measurements of surface reflectance and atmospheric parameters. For the GF5-02 AHSI, the shifts in the central wavelength of the Visible and Near-Infrared (VNIR) band is 0.117 nm, while the shifts in the Full Width at Half Maximum (FWHM) is 0.02 nm. In the Short-Wave Infrared (SWIR) band, these values are 0.25 nm for the central wavelength and 0.04 nm for the FWHM. The results demonstrate that the applied method is effective for on-orbit spectral calibration for GF5-02 AHSI.

Interband radiometric calibration from the mid-infrared to visible bands in the ocean specular region is an effective way to calibrate on-orbit remote sensing sensors. It assumes that the referenced band has highly accurate radiance and that the interband radiometric relationship can be obtained in the ocean specular region. Most current research employs only the radiative transfer (RT) equation to derive interband radiometric relationships. However, two variables—water-leaving radiance and whitecaps—are challenging to obtain yet crucial for radiative transfer calculations. Typically, water-leaving radiance is assigned a fixed value since empirical data, whereas whitecaps are estimated via the wind speed alone. These assumptions make the uncertainties of the calibrated bands large and different from those of real satellite-measured data, reducing the reliability of the interband relationship between the reference and calibrated bands and limiting the application of the interband radiometric calibration method. To address this issue, this study proposed a novel interband radiometric calibration method called coupled deep neural networks and radiative transfer (CDR), which integrates radiative transfer and a deep neural network (DNN) to provide a reliable relationship between referenced and to be calibrated bands without accurate water-leaving radiance and whitecaps. For the four visible bands of FY-3D/MERSI-II, the relative errors were found to be 2.12%, 4.62%, 1.89%, and 4.02%, respectively. Uncertainty analysis identified the referenced band as the largest uncertainty source, followed by chlorophyll concentration, polarization effects, and aerosol loading. The CDR algorithm can be used to calibrate historical long-term satellite data without additional measurements.

The urban land surface temperature (LST) is very important in urban development, changes and local climate in the city, etc. Many methods have been proposed to inverse LST from satellite remotely sensed data. In this study, an inversion method of combining deep learning and physical model was proposed to estimate the urban surface temperature from CHINESE SDGSAT-1 satellite, i.e., an improved temperature and emissivity separation (TES) algorithm based on sky view factor (SVF). Finally, two data sets were selected to evaluate the accuracy of the proposed algorithm. Results show that the root mean squared errors (RMSEs) using the proposed algorithm are approximately 0.39K for LST and 0.023 for emissivity, respectively. The proposed algorithm is applied to inverse the LST/LSE of Beijing and Wuhan, China. Compared with Landsat-8 satellite products, the proposed algorithm has consistent accuracy. The cross-validated RMSEs with the Landsat-8 surface temperature product in Wuhan and Beijing were 2.18 K and 1.11 K, respectively.

Semantic segmentation is an essential technique in remote sensing. Until recently, most related research has focused primarily on advancing semantic segmentation models based on monomodal imagery, and less attention has been given to models that utilize multimodal remote sensing data. Moreover, most current multimodal approaches consider only limited bimodal situations and cannot simultaneously utilize three or more modalities. The increase in expensive computational costs associated with previous feature fusion paradigms hinders their application in broader cases. How to design a unified method to cover a wide variety of quantity-agnostic modalities for multimodal semantic segmentation remains unsolved issues. To address the aforementioned challenges, this study explores a feasible way and proposes a cost-effective multimodal sensing semantic segmentation model (MultiSenseSeg). MultiSenseSeg employs multiple lightweight modality-specific experts (MSEs), an adaptive multimodal matching (AMM) module, and a single feature extraction pipeline to efficiently model intra- and inter-modal relationships. Benefiting from these designs, the proposed MultiSenseSeg can serve as a unified multimodal model capable of addressing both monomodal and bimodal cases and readily extrapolating to scenarios with more modalities, thereby achieving semantic segmentation of arbitrary quantities of multimodal data. To evaluate the performance of our method, we select several state-of-the-art (SOTA) semantic segmentation models from the past three years and conduct extensive experiments on two public multimodal datasets. The results show that MultiSenseSeg can not only achieve higher accuracy but also exhibits user-friendly modality extrapolation, allowing end-to-end training for consumer-grade users based on limited hardware resources. The model’s code will be available at https://github.com/W-qp/MultiSenseSeg.

The separation of land surface temperature (LST) and land surface emissivity (LSE) is an ill-posed problem in thermal infrared (TIR) remote sensing. By building a new observation matrix to compress the LSE unknows and a dictionary training method to reconstruct complete LSE spectra, a new dictionary-based sparse representation for emissivity (DSRE) method has been proposed to retrieve LST and LSE from the atmospherically corrected hyperspectral TIR data. The proposed method fully utilizes the sparsity of compressed sensing and the empirical knowledge of the trained emissivity dictionary. The sensitivity analysis shows that the modeling accuracies of the proposed method are 0.215Kand 0.0060 for LST and LSE, respectively. Even with the instrument noise of 0.3Kand the uncertainties in atmospheric transmittance, atmospheric upwelling, and downwelling radiance of 10 %, the retrieval accuracies are 0.811Kfor LST and 0.0241 for LSE, respectively. Then a field experiment was conducted to validate the proposed method, and a comparison was executed to three published methods, including ASTER temperature-emissivity separation (ASTERTES), linear spectral emissivity constraint TES (LSECTES), and iterative spectrally smooth TES (ISSTES). The accuracies of retrieved LST and spectral LSE are 1.41K/ 0.009, 2.57K/ 0.071, 1.59K/ 0.038, and 2.00K/ 0.077 for DSRE, ASTERTES, LSECTES, and ISSTES. In contrast to the three published methods, our proposed method is more accurate and effective than other published methods. Especially in the atmospheric absorption band, the proposed method has a strong anti-noise capability to the residuals of environmental downwelling radiance.

High resolution thermal infrared satellites can provide important observations for applications in various fields. For mine detection application, high resolution data can better resolve mineralogic boundaries. In this paper, an emissivity image simulation method is proposed for a high resolution thermal infrared satellite concept; and the simulated images can be used for mineral recognition and classification algorithm development using such instrument specification. Using Worldview-3 imagery as the simulation data source, the emissivity image has been generated based on a linear mixing model. In addition, accuracy analysis is conducted through comparison between simulated WorldView-3 images and the actual one. Small relative errors in every WorldView-3 band show it is feasible to use our proposed method for image simulation.

In this paper, radiometric cross-calibration of Ziyuan-3 Satellite was conducted at the homogeneous field test site in August 2015, using GF1 satellite and Landsat-8. The test site was chosen at Dunhuang site, which has been used for vicarious calibration for Chinese Satellite since 1990s. The radiometric cross-calibration of ZY3, GF1 and Landsat-8 was performed by the simultaneously observation of the three satellites on August 15, 2015. In this study, three calibration coefficients from various methods were compared and analyzed. This paper reveals that the cross-calibration of the ZY3 satellite using GF1 PMS sensor is better than using GF1 WFV sensor. The best result is the cross-calibration coefficient using Landsat-8 OLI, which is consistent with the calibration coefficient of the reflectance-based vicarious calibration method in Dunhuang. The validation results show the accuracy of the cross-calibration for ZY3 sensor using GF1 PMS and Landsat-8 OLI is less than 4% and 2%,respectively. This study can supply reference to the radiometric cross-calibration of Chinese satellite senor, and the results suggest that the radiometric cross-calibration coefficients can be useful for maintaining the quantitative application.

The blind areas of two adjacent aerial images are usually large in Urban Remote Sensing, in order to reconstruct three-dimensional (3D) information of the building in these blind areas, a new modified SFS algorithm based on stereo images is proposed. This SFS algorithm is utilized to reconstruct object surface gradient information (relative elevation) of buildings in the overlapping areas based on the grey information of the image. The absolute elevation information could be gained by the stereo images which had been already finished relative orientation and absolute orientation. The fitting transformational relationship is set up between the relative and absolute elevation information of the characteristic ground points in these overlapping areas with Least Square Method. The building 3D information in non-overlapping domain can be reconstructed by this transformational relation. From the experiment results, it is found that the stereo models reconstructed by the modified SFS algorithm based on stereo images are better and have lower distortions than those reconstructed by general SFS algorithm.

The feasibility of polarization remote sensing (RS) system is briefly analyzed for unmanned aerial vehicle (UAV) in China. With system integrated method, polarization RS system for UAV is designed based on long-term polarization theoretical research and ground experiments. The system structure and workflow are introduced, with software and hardware. With successful joint debugging on the ground, the system can satisfy operational requirements for further aviation. It compensates the limitations in types and performance for UAV remote sensing system in China.