Land surface emissivity (LSE) can reflect the unique spectral characteristics of land surfaces, which holds significant research value and promising applications. However, due to the limitations of spaceborne hyperspectral thermal infrared (TIR) sensors, there is currently a lack of high spatial resolution LSE imagery. Therefore, this study proposes a methodology based on machine learning and mathematical statistical approaches to extend multispectral emissivity values to hyperspectral LSE spectrums. Furthermore, the Spectral Angle Mapper (SAM) method was used for target identification. The results indicated that the proposed method can estimate hyperspectral LSE curves in high accuracy. Moreover, the target ground class can be identified accurately by matching the spectral angle with the measured LSE curve. This has crucial implications for applications such as land cover classification, mineral identification, geological exploration and soil property research.

Land surface temperature (LST) and emissivity (LSE) are the crucial parameters for thermal infrared (TIR) remote sensing. However, the coupling of the two parameters presents a challenge to achieving high-accuracy retrieval, particularly for sensors with only one or two TIR channels. Following the launch of Landsat 9, there has been a rapid increase in demand for methods to accurately estimate LSE and LST for sensors with high spatial resolution but limited TIR channels. Therefore, this article proposes a two-step framework to retrieve LSE and LST for Landsat 9 only using data of its own. First, the data in visible-to-near-infrared (VNIR) to short-wave infrared (SWIR) channels of Landsat 9 were used to retrieve LSEs based on a machine learning method. Subsequently, the split-window (SW) method was employed to retrieve LST based on the estimated LSEs. As a result, the retrieved LSE exhibits high accuracy across the cross and direct validation, with RMSEs all below 0.01 for the two TIR channels. For LST, the retrieved result was validated by the existing products and in situ LSTs from surface radiation budget (SURFRAD), demonstrating excellent accuracies, with RMSE of 1.86 K, which is superior to the LST product of Landsat 9, with RMSE of 2.14 K. Therefore, the proposed framework is feasible for LSE and LST retrieval without support of auxiliary data from other origins, which is of great significance for the sensors with limited TIR channels to produce accurate LSE and LST products.

As one of the key parameters in the physics of land-surface processes, land surface temperature (LST) plays an important role in many fields, such as urban heat island effect, forest fire monitoring, drought monitoring and so on. Thermal infrared remote sensing is the main way to obtain LST in large scale. For the sensors with one TIR channel, the single channel methods are mostly used. These methods need to know atmospheric parameters and emissivity. However, the accuracy of atmospheric correction is different to guarantee in many cases, which limited the applicability and accuracy of these methods. Therefore, based on the channel correlation hypothesis, a single channel algorithm without atmospheric correction is proposed in this paper. The feasibility of this method is preliminarily verified by simulated and satellite data.

The nightlight data can objectively represent the regional scope of human activities at night, so as to reflect the economic activities at night. Based on SDGSAT-1 satellite data, this paper analyzes the urban living space factor index and urban regional overnight economic activity index of Beijing before and after the lifting of COVID-19 prevention and control policy in 2022. The results show that before and after the lifting of COVID-19 prevention and control policy, There was a significant correlation between the change of night light index and human activities in the two periods, and the correlation was more than 63%. There are obvious differences in the economic activity index of urban areas at night, which reflects that the ULSF index calculated based on high-resolution luminous images and POI data is feasible and reliable to analyze the changes of urban living space at night.

Landsat data are the important sources for the inversion of land surface temperature (LST) with high spatial resolution. As an important parameter of LST inversion, the land surface emissivity (LSE) of Landsat TIR channels is usually inverted by the semi-empirical method, which has certain limitations. With the launch of Landsat 9, the requirement for LSE became more urgent. Therefore, this paper proposed a new retrieval method to estimate LSE with high spatial resolution for Landsat. Reflectance of Landsat VNIR-SWIR channels were used for the LSE inversion by Gradient Boost Regression Tree (GBRT) machine learning method. In order to evaluate the accuracy of the estimated results, the LSEs were compared with those estimated by NDVI threshold method and the field measurement data. The results showed the LSEs estimated by GBRT model were consistent with the measured data. So it demonstrated that this method was feasible to estimated LSE for Landsat.

Land Surface Emissivity (LSE) is an important parameter in thermal infrared remote sensing, which is of great significance to temperature inversion. In this study, the Gradient Boost Regression Tree (GBRT) was proposed to directly retrieve LSEs of MODIS thermal infrared channels 29$(8.4-8.7\ \mu \mathrm{m}), 31(10.78-11.28\ \mu \mathrm{m})$, and 32 ($11.77-12.27\ \mu \mathrm{m}$) from the visible and near infrared (VNIR) data. We selected the variables related with LSE, including reflectivity, view zenith, solar zenith, land surface type, vegetation index (EVI), Normalized Difference Water Index (NDWI) and Leaf Area Index (LAI). The results of the test set showed that RMSEs of the estimated LSEs were 0.013 in channel 29, 0.005 in 31 and 0.004 in 32, which were more accurate than existing methods. Eight regions with different ground features were also selected to further evaluate the applicability of the model. In most areas, the RMSEs were below 0.015 in channel 29, below 0.005 in channel 31 and 32. In addition, the spatial distributions of the estimated LSEs and those extracted from MYD11B1 and MYD21A1D in H19V08 were compared, which were also reasonable. In general, it is feasible to use the selected variables with the GBRT model to directly retrieve the LSEs.

Bit allocation plays an important role in rate control for it determines the calculation of other parameters of the rate control model. For surveillance videos, however, the bit distribution analysis shows that the foreground parts should get more bits and be encoded in higher quality than background parts. By utilizing the background and foreground information (BFI) provided by surveillance videos, this paper proposes a background-foreground information based bit allocation algorithm (BFIBA). The main idea of BFIBA is to classify a largest coding unit (LCU) into a foreground LCU or a background LCU according to BFI and then allocate bits in both frame level and LCU level and adjust the quantization parameter by using the information provided by the classification. Experimental results show that BFIBA maintains rate control performance, obtaining a rate control error of 0.1% and the foreground coding quality increases by 0.36dB, 0.76dB and 0.82dB in PSNR compared with the three existing bit allocation algorithms in HEVC rate control.