Article
Environmental Sciences
Kumari Preety, Anup K. Prasad, Atul K. Varma, Hesham El-Askary
Summary: Publicly available digital elevation models have greatly influenced the quantification of landscape characteristics. This study examines the accuracy of six major public domain satellite-derived DEMs and finds that applying DGPS correction can improve accuracy significantly.
Article
Geochemistry & Geophysics
Binbin Li, Huan Xie, Xiaohua Tong, Hong Tang, Shijie Liu
Summary: This article proposes a DEM correction model using ICESat-2 data, which verifies DEM elevation errors in ICESat-2 coverage areas and evaluates error sources of global-scale DEMs. A regression model is then constructed to correct DEMs in areas without ICESat-2 coverage. Validation experiments show that the proposed model is suitable for global-scale DEM elevation correction.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Environmental Sciences
Zhoujin Wang, Lichun Sui, Shiqi Zhang
Summary: The land surface temperature (LST) images obtained by thermal infrared remote sensing sensors play an important role in various research fields. However, the low spatial resolution of these images is a limitation. To address this issue, the authors propose a downscaling model based on the Sentinel-3 satellite and ASTER GDEM images. Through comparisons with other models, it is demonstrated that the proposed model can improve the spatial resolution and visual effects of the images while maintaining high temporal resolution. The influence of terrain and land cover on the image data is also discussed.
Article
Geosciences, Multidisciplinary
Vladimir Tabunshchik, Roman Gorbunov, Tatiana Gorbunova, Cam Nhung Pham, Aleksandra Klyuchkina
Summary: Since the end of the 20th century, the use of geographic information systems and digital elevation models has improved morphometric analysis of river basins. In this study, we compared the accuracy of different digital elevation models for river basins in the northwestern slope of the Crimean Mountains. The Copernicus DEM and ALOS World 3D datasets showed the smallest errors, and we used the Copernicus DEM dataset to model the river basins in the study area.
FRONTIERS IN EARTH SCIENCE
(2023)
Article
Environmental Sciences
Haijiao Han, Qiming Zeng, Jian Jiao
Summary: Digital elevation models (DEMs) are fundamental data for scientific and engineering research, and the quality of DEMs can impact local applications. Studies show that TanDEM-X DEM has the best overall quality followed by SRTM, while ASTER GDEM has the lowest quality. The relationship between DEM quality and terrain factors, as well as the creation process of DEMs, is significant for further applications.
Article
Engineering, Electrical & Electronic
Yanli Zhang, Yan Pang, Dudu Cui, Yupeng Ma, Linhong Chen
Summary: The study validated the elevation accuracy of ICESat-2/ATL06 data for monitoring changes in glacier thickness using field surveys with CORS and UAVs. The experiments demonstrated that the data have high vertical and horizontal accuracy, with RMSE values of 0.0846 m from CORS and 0.1517 m from UAVs, and positioning accuracy is influenced by terrain slope.
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
(2021)
Article
Geography, Physical
Tong Han, Huaguo Zhang, Wenting Cao, Chengfeng Le, Chen Wang, Xinke Yang, Yunhan Ma, Dongling Li, Juan Wang, Xiulin Lou
Summary: Accurate bathymetric mapping of shallow oceanic islands is difficult due to limited availability of high-quality optical remote sensing images. In this study, a costefficient method based on a quadratic polynomial ratio model (QPRM) of massive active-passive remote sensing data was proposed. The results showed that the QPRM method has a slight advantage in terms of accuracy and efficiency compared to the classical linear ratio model (CLRM) and multiple fitting methods, making it suitable for bathymetric mapping of oceanic islands and reefs worldwide.
ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING
(2023)
Article
Geosciences, Multidisciplinary
Farah Abdelouhed, Algouti Ahmed, Algouti Abdellah, Ifkirne Mohammed, Ourhzif Zouhair
Summary: This study focused on using remote sensing techniques and GIS to automatically extract lineaments in southeast Morocco, and found that the lineaments were mainly concentrated in the northeastern part of the Central High Atlas and the Precambrian bedrock of Saghro, with a high number of lineaments ranging in length from a few meters to 7.5 km.
Article
Environmental Sciences
Zhiyu Zhang, Xinyuan Liu, Yue Ma, Nan Xu, Wenhao Zhang, Song Li
Summary: This paper proposes an effective algorithm to extract signal photons from weak beam data of ICESat-2 in mountainous areas by utilizing strong beam data and the relationship between along-track slope and noise level. Through this method, a higher detection rate of signal photons is achieved, improving the overall performance compared to other algorithms. The algorithm shows potential in assessing elevation accuracy achieved by ICESat-2, estimating along-track and cross-track slope, and obtaining ground control points for stereo-mapping satellites in mountainous regions.
Article
Environmental Sciences
Xin Zhang, Baorong Xie, Shijie Liu, Xiaohua Tong, Rongli Ding, Huan Xie, Zhonghua Hong
Summary: This paper proposes a two-step block adjustment approach for improving the accuracy and consistency of Digital Surface Models (DSMs). By using high-accuracy ICESat-2 laser altimetry data as elevation control, this method efficiently enhances the elevation accuracy of DSMs acquired by different sensor types.
Article
Multidisciplinary Sciences
Baojin Han, Min Hu
Summary: Due to smaller datasets and unbalanced sample distribution, the current facial emotion datasets have limited effectiveness in facial expression recognition. Traditional data augmentation methods are insufficient in solving the problem of high similarity among generated images. StarGAN V2 can generate diverse styles of images across multiple domains, but it has drawbacks like distorted mouths and blurry facial expression images. To address these issues, we improved StarGAN V2 by applying SENet to its generator, allowing it to focus on important regions for clearer and distinguishable symmetrical expression images. We also customized the hinge loss function to enhance the quality of generated images and conducted experiments on CK+ and MMI datasets to demonstrate the superiority of our improved model over state-of-the-art methods.
Article
Environmental Sciences
Yi Li, Haiqiang Fu, Jianjun Zhu, Kefu Wu, Panfeng Yang, Li Wang, Shijuan Gao
Summary: The rapid development of the SRTM digital elevation model has been witnessed in the past decade, but it has poor elevation accuracy in forested areas. The latest ICESat-2 data provides an opportunity to correct the elevation error of the SRTM DEM, and a method based on spatial interpolation has been proposed and tested in a forested area in Massachusetts, USA, achieving significant improvements in accuracy.
Article
Environmental Sciences
Tao Wang, Yong Fang, Shuangcheng Zhang, Bincai Cao, Zhenlei Wang
Summary: This study investigates the analysis and calibration of relative systematic biases in ICESat-2 laser data. The crossover method is used to calibrate the relative bias between different beams, improving the accuracy and consistency of the data.
Article
Geography, Physical
Jieying Lao, Cheng Wang, Sheng Nie, Xiaohuan Xi, Hui Long, Baokun Feng, Zijia Wang
Summary: A new denoising method is proposed to adaptively extract signal photons in spaceborne photon-counting LiDAR. The method includes two steps: the first-step denoising process is completed based on photons' curvature feature, and the second-step eliminates local dense noise photons. Experimental results show that the proposed algorithm can accurately identify signal photons in different surface types and topographies, providing more reasonable and reliable data for sustainable urban development.
INTERNATIONAL JOURNAL OF DIGITAL EARTH
(2023)
Article
Chemistry, Analytical
Linyu Gu, Dazhao Fan, Song Ji, Zhihui Gong, Dongzi Li, Yang Dong
Summary: In this study, a new method for extracting signal photons from laser altimetry data is proposed. The method combines slope and elevation information from optical stereo images and adaptively adjusts the neighborhood search direction based on the spatial distribution of noise and signal density. The algorithm performs better than the existing ATL08 algorithm in steep slope and low SNR regions, and can accurately extract continuous and reliable surface signals in different terrains and land cover types.