Qisheng Wang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Qisheng Wang has a total of 25 co-authors. Here are the top co-authors by number of shared publications:
- Shuanggen Jin
Henan Polytechnic University
4 shared publications Jiaru Zhu
3 shared publications- Dengshan Huang
College of Civil Engineering, Xiangtan University, Xiangtan 411105, China
2 shared publications Feng Hu
Guizhou Phosphate Chemical Group Company Ltd., Guiyang, China
2 shared publications- Yulin Tang
College of Civil Engineering, Xiangtan University, Xiangtan 411105, China
2 shared publications - Zhen Zhou
School of Civil Engineering, Xiangtan University, Xiangtan, 411105, People’s Republic of China
2 shared publications
Latest Publications
- Open Access
Fine-Grained Natural Grassland Mapping for Sustainable Land Monitoring in Urban–Rural Landscapes Using Foundation Model Embeddings
Fine-grained information on natural grasslands is important for monitoring ecological land patterns and supporting sustainable land management in heterogeneous urban–rural landscapes. However, natural grasslands in subtropical southern China are highly fragmented and frequently embedded within forests, croplands, settlements, and other land-cover types, making their fine-grained identification challenging using conventional land-cover products and single-season remote sensing observations. This study focuses on Hunan Province and constructs a five-class dataset using field-survey plots independently collected during our field campaign, together with public land-cover products for spatially reliable sample screening. Seasonal–phenological features derived from spring, summer, and autumn Sentinel-2 observations, topographic information from DEM, and AlphaEarth Embedding were integrated into a proposed Grassland Mapping Network (GMapNet). GMapNet employs a Transformer to model dependencies among feature groups, Feature Group Attention (FGA) to adaptively aggregate seasonal and environmental information, and a Class Prototype Similarity Head (CPSH) for prototype-based fine-grained discrimination. GMapNet achieved an overall accuracy of 88.58 ± 0.98%, a macro-F1 of 88.46 ± 1.05%, and a Kappa coefficient of 85.72 ± 1.23%, outperforming all compared classification methods. Ablation experiments further confirmed the effectiveness of FGA and CPSH. Removing AlphaEarth Embedding reduced the overall accuracy to 76.58 ± 1.72%, demonstrating that foundation-model embeddings provide substantial complementary information beyond conventional seasonal–phenological–topographic features. The resulting maps for four representative cities displayed pronounced spatial heterogeneity, with predicted natural grasslands distributed mainly as fragmented patches and terrain-associated belts within complex land-cover mosaics. These maps provide finer thematic detail than broad grassland/open-vegetation categories and may inform future ecological monitoring and conservation planning.
Open AccessFine-Grained Natural Grassland Mapping for Sustainable Land Monitoring in Urban–Rural Landscapes Using Foundation Model Embeddings
Fine-grained information on natural grasslands is important for monitoring ecological land patterns and supporting sustainable land management in heterogeneous urban–rural landscapes. However, natural grasslands in subtropical southern China are highly fragmented and frequently embedded within forests, croplands, settlements, and other land-cover types, making their fine-grained identification challenging using conventional land-cover products and single-season remote sensing observations. This study focuses on Hunan Province and constructs a five-class dataset using field-survey plots independently collected during our field campaign, together with public land-cover products for spatially reliable sample screening. Seasonal–phenological features derived from spring, summer, and autumn Sentinel-2 observations, topographic information from DEM, and AlphaEarth Embedding were integrated into a proposed Grassland Mapping Network (GMapNet). GMapNet employs a Transformer to model dependencies among feature groups, Feature Group Attention (FGA) to adaptively aggregate seasonal and environmental information, and a Class Prototype Similarity Head (CPSH) for prototype-based fine-grained discrimination. GMapNet achieved an overall accuracy of 88.58 ± 0.98%, a macro-F1 of 88.46 ± 1.05%, and a Kappa coefficient of 85.72 ± 1.23%, outperforming all compared classification methods. Ablation experiments further confirmed the effectiveness of FGA and CPSH. Removing AlphaEarth Embedding reduced the overall accuracy to 76.58 ± 1.72%, demonstrating that foundation-model embeddings provide substantial complementary information beyond conventional seasonal–phenological–topographic features. The resulting maps for four representative cities displayed pronounced spatial heterogeneity, with predicted natural grasslands distributed mainly as fragmented patches and terrain-associated belts within complex land-cover mosaics. These maps provide finer thematic detail than broad grassland/open-vegetation categories and may inform future ecological monitoring and conservation planning.
- 17 August 2026
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Space Physics
- Vol. 131 (8)
A High‐Performance Method for Generating Ionospheric Tomographic Projection Matrix Using Spherical Triangles
Projection‐matrix generation is an essential prerequisite for ionospheric tomographic inversion, and its computational speed determines the practicality of high‐resolution tomography. We developed and tested an algorithm that combines a fast ray‐voxel intersection test with the spherical‐triangle method. The algorithm was evaluated against the conventional formulation, the parametric‐equation approach, and the segmentation intersection and voxel tracking (SIVT) approach using two European GNSS epochs and voxel grids at different resolutions. Within these tested configurations, the spherical‐triangle method alone performed well for coarse grid configurations, whereas the combined “fast intersection test scheme II + spherical‐triangle method” gave the shortest runtime for the finest grid resolution. At a resolution of 0.1 (longitude) × 0.1 (latitude) × 1 km (altitude), the combined method completed the computations in approximately 3.88 and 8.45 s for 9,225 and 17,081 valid rays, respectively, corresponding to a 40%–50% reduction relative to using parametric equations alone in the same test setting. With five‐thread parallelization, the processing time was further reduced to 1.31 and 2.62 s. These results indicate that the proposed workflow can reduce the cost of projection‐matrix generation for the tested ionospheric tomography configurations without altering the numerical results.
- 17 August 2026
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Space Physics
- Vol. 131 (8)
A High‐Performance Method for Generating Ionospheric Tomographic Projection Matrix Using Spherical Triangles
Projection‐matrix generation is an essential prerequisite for ionospheric tomographic inversion, and its computational speed determines the practicality of high‐resolution tomography. We developed and tested an algorithm that combines a fast ray‐voxel intersection test with the spherical‐triangle method. The algorithm was evaluated against the conventional formulation, the parametric‐equation approach, and the segmentation intersection and voxel tracking (SIVT) approach using two European GNSS epochs and voxel grids at different resolutions. Within these tested configurations, the spherical‐triangle method alone performed well for coarse grid configurations, whereas the combined “fast intersection test scheme II + spherical‐triangle method” gave the shortest runtime for the finest grid resolution. At a resolution of 0.1 (longitude) × 0.1 (latitude) × 1 km (altitude), the combined method completed the computations in approximately 3.88 and 8.45 s for 9,225 and 17,081 valid rays, respectively, corresponding to a 40%–50% reduction relative to using parametric equations alone in the same test setting. With five‐thread parallelization, the processing time was further reduced to 1.31 and 2.62 s. These results indicate that the proposed workflow can reduce the cost of projection‐matrix generation for the tested ionospheric tomography configurations without altering the numerical results.
- 1 April 2026
- Published by IOP Publishing in Engineering Research Express
- Vol. 8 (7) , 075230
An improved method for single-station receiver DCB estimation based on a binary search algorithm
Differential Code Bias (DCB), also referred to as hardware delay, directly affects the accuracy of positioning, timing, ionospheric research, and the reliability of multi-system integration. Accurate DCB estimation is critical for these applications. Currently, two primary approaches are used to estimate DCB: 1) correcting the Total Electron Content (TEC) using empirical models and then solving for DCB and 2) developing an ionospheric model in which both the Vertical Total Electron Content (VTEC) and DCB are estimated simultaneously as unknown parameters. The minimum standard deviation over all lines of sight method provides a straightforward approach for estimating DCB. This method neither requires empirical models nor ionospheric modeling, making it computationally efficient; however, its performance depends on the search step size. In this study, a modified method based on binary search is proposed to enhance computational efficiency. Using GPS data from IGS stations in January 2014 and January 2020, the DCB of the receiver was estimated via both the original and the modified methods. Computational efficiency was evaluated, and the results were compared with IGS products. The findings indicate that the DCB estimated by the modified method exhibits an absolute mean deviation and RMS of 1.02 ns and 1.11 ns in 2014 and 0.80 ns and 0.85 ns in 2020, respectively, compared to the IGS products. The results from the modified method show a high degree of consistency with those from the original method, with a correlation coefficient close to 1. In terms of computational efficiency, the modified method is 6–9 times faster than the original method with a search step size of 0.1 ns and 45–116 times faster than the original method with a search step size of 0.01 ns. Experimental results confirm the accuracy and effectiveness of the modified method.
- 1 April 2026
- Published by IOP Publishing in Engineering Research Express
- Vol. 8 (7) , 075230
An improved method for single-station receiver DCB estimation based on a binary search algorithm
Differential Code Bias (DCB), also referred to as hardware delay, directly affects the accuracy of positioning, timing, ionospheric research, and the reliability of multi-system integration. Accurate DCB estimation is critical for these applications. Currently, two primary approaches are used to estimate DCB: 1) correcting the Total Electron Content (TEC) using empirical models and then solving for DCB and 2) developing an ionospheric model in which both the Vertical Total Electron Content (VTEC) and DCB are estimated simultaneously as unknown parameters. The minimum standard deviation over all lines of sight method provides a straightforward approach for estimating DCB. This method neither requires empirical models nor ionospheric modeling, making it computationally efficient; however, its performance depends on the search step size. In this study, a modified method based on binary search is proposed to enhance computational efficiency. Using GPS data from IGS stations in January 2014 and January 2020, the DCB of the receiver was estimated via both the original and the modified methods. Computational efficiency was evaluated, and the results were compared with IGS products. The findings indicate that the DCB estimated by the modified method exhibits an absolute mean deviation and RMS of 1.02 ns and 1.11 ns in 2014 and 0.80 ns and 0.85 ns in 2020, respectively, compared to the IGS products. The results from the modified method show a high degree of consistency with those from the original method, with a correlation coefficient close to 1. In terms of computational efficiency, the modified method is 6–9 times faster than the original method with a search step size of 0.1 ns and 45–116 times faster than the original method with a search step size of 0.01 ns. Experimental results confirm the accuracy and effectiveness of the modified method.
- 29 December 2025
- Published by IOP Publishing in Engineering Research Express
- Vol. 7 (4) , 0452h5
YOLOv8s-LS: a lightweight road crack detection model based on UAV imagery with multi-scale optimization
This study introduces YOLOv8s-LS (Lightweight and Scalable), a novel lightweight target detection model based on YOLOv8s. It is designed to meet the dual requirements of real-time performance and high accuracy in road crack detection under resource-constrained environments. The model integrates multi-scale ghost convolution (MSGConv), a multi-scale dilated local attention C2f module (C2f_MDLA), multi-channel non-local spatial attention (MNSA), and dimension-preserving separable convolution (MODSConv). These modules allow efficient extraction of multi-scale features while significantly reducing computational complexity. This innovative ‘lightweight backbone + fine attention’ paradigm advances beyond recent YOLOv8 improvements. It leverages synergistic module design tailored to the elongated morphologies and scale variations of cracks in UAV imagery. Comparative experiments on the UAV-PDD2023 dataset show that YOLOv8s-LS improves mAP50 and mAP50–95 by 9.0% and 17.3%, respectively. It also reduces computational cost by 53.1% compared to YOLOv8s. On the RDD2020 dataset, mAP50 increases by 2.4 percentage points. Transferring the optimizations to YOLOv8n yields a 6.1 percentage point increase in mAP50. Model size and power consumption are reduced by 46.8% and 30.1%, respectively. Multi-hardware benchmarks further demonstrate the transfer’s efficacy: a 52.9% reduction in GFLOPs translates to 39%–55% FPS gains across platforms (RTX 3060 GPU, i7-12700 CPU, and Jetson Nano). These results validate YOLOv8s-LS’s superior detection performance and lightweight deployment advantages in complex scenarios. The model offers an efficient solution for UAV-based road disease inspection and civil engineering applications.
- 29 December 2025
- Published by IOP Publishing in Engineering Research Express
- Vol. 7 (4) , 0452h5
YOLOv8s-LS: a lightweight road crack detection model based on UAV imagery with multi-scale optimization
This study introduces YOLOv8s-LS (Lightweight and Scalable), a novel lightweight target detection model based on YOLOv8s. It is designed to meet the dual requirements of real-time performance and high accuracy in road crack detection under resource-constrained environments. The model integrates multi-scale ghost convolution (MSGConv), a multi-scale dilated local attention C2f module (C2f_MDLA), multi-channel non-local spatial attention (MNSA), and dimension-preserving separable convolution (MODSConv). These modules allow efficient extraction of multi-scale features while significantly reducing computational complexity. This innovative ‘lightweight backbone + fine attention’ paradigm advances beyond recent YOLOv8 improvements. It leverages synergistic module design tailored to the elongated morphologies and scale variations of cracks in UAV imagery. Comparative experiments on the UAV-PDD2023 dataset show that YOLOv8s-LS improves mAP50 and mAP50–95 by 9.0% and 17.3%, respectively. It also reduces computational cost by 53.1% compared to YOLOv8s. On the RDD2020 dataset, mAP50 increases by 2.4 percentage points. Transferring the optimizations to YOLOv8n yields a 6.1 percentage point increase in mAP50. Model size and power consumption are reduced by 46.8% and 30.1%, respectively. Multi-hardware benchmarks further demonstrate the transfer’s efficacy: a 52.9% reduction in GFLOPs translates to 39%–55% FPS gains across platforms (RTX 3060 GPU, i7-12700 CPU, and Jetson Nano). These results validate YOLOv8s-LS’s superior detection performance and lightweight deployment advantages in complex scenarios. The model offers an efficient solution for UAV-based road disease inspection and civil engineering applications.
- 26 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 64 (01962892) , 1-10
An Independent Estimation Method for GNSS Code Bias Based on Direct Processing Strategy of Single Station TEC
Global navigation satellite system (GNSS) code biases are a significant source of error in ionospheric Total electron content (TEC) modeling and inversion, and they need to be accurately determined. Since the code bias parameters and ionospheric TEC parameters are coupled together in the observation equations, they cannot be directly separated. Therefore, the estimation of code biases must take into account the processing strategy for the TEC parameters. The conventional methods mainly involve function fitting or using existing information for elimination, relying on function models and Global ionosphere maps (GIM), and cannot achieve direct estimation. Based on existing research, this paper attempts to directly estimate the combined code bias values for satellites and receivers at individual stations by simplifying the processing of TEC using single-station observation data. Then, by combining regional or global stations, the code bias values for satellites and receivers can be further obtained, achieving a new method for code bias estimation. Two datasets from the first month of 2014 (a high solar activity years) and 2020 (a low solar activity years) were selected for experimental analysis. The RMS of the estimated TEC for all stations relative to CODE’s GIM was mostly within 3 TECu, with mean values of approximately 4 TECu and 2 TECu, respectively. Using the Differential code bias (DCB) products provided by CODE as a reference, the RMS of the estimated DCBs for GPS and GLONASS satellites in 2014 were 0.14 and 0.42ns, and that in 2020 were 0.07 and 0.18 ns, showing good consistency and validating the effectiveness of the method.
- 26 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 64 (01962892) , 1-10
An Independent Estimation Method for GNSS Code Bias Based on Direct Processing Strategy of Single Station TEC
Global navigation satellite system (GNSS) code biases are a significant source of error in ionospheric Total electron content (TEC) modeling and inversion, and they need to be accurately determined. Since the code bias parameters and ionospheric TEC parameters are coupled together in the observation equations, they cannot be directly separated. Therefore, the estimation of code biases must take into account the processing strategy for the TEC parameters. The conventional methods mainly involve function fitting or using existing information for elimination, relying on function models and Global ionosphere maps (GIM), and cannot achieve direct estimation. Based on existing research, this paper attempts to directly estimate the combined code bias values for satellites and receivers at individual stations by simplifying the processing of TEC using single-station observation data. Then, by combining regional or global stations, the code bias values for satellites and receivers can be further obtained, achieving a new method for code bias estimation. Two datasets from the first month of 2014 (a high solar activity years) and 2020 (a low solar activity years) were selected for experimental analysis. The RMS of the estimated TEC for all stations relative to CODE’s GIM was mostly within 3 TECu, with mean values of approximately 4 TECu and 2 TECu, respectively. Using the Differential code bias (DCB) products provided by CODE as a reference, the RMS of the estimated DCBs for GPS and GLONASS satellites in 2014 were 0.14 and 0.42ns, and that in 2020 were 0.07 and 0.18 ns, showing good consistency and validating the effectiveness of the method.
- 13 September 2024
- Published by Pleiades Publishing Ltd in Numerical Analysis and Applications
- Vol. 17 (3) , 215-224
Partial Multivariate Errors-in-Variables Model and Its Application in Settlement Monitoring
Aiming at the problem that the coefficient matrix of multivariate errors-in-variables (MEIV) model contains constant columns, the MEIV model is extended to Partial multivariate errors-in-variables (P-MEIV), and the new algorithm of P-MEIV model is proposed based on the principle of Partial errors-in-variables (PEIV) model and indirect adjustment. The algorithm is simple and easy to implement. An example of coordinate transformation is used for verifying, and the results are compared with the existing MEIV model algorithm, which shows the effectiveness of the proposed algorithm. Finally, the P-MEIV algorithm is applied to the multi-point grey model (MGM(1,N)) of settlement monitoring. The results show that the P-MEIV model proposed in this paper can better consider the influence of monitoring point errors, and the estimated results are in good agreement with the actual situation.
- 13 September 2024
- Published by Pleiades Publishing Ltd in Numerical Analysis and Applications
- Vol. 17 (3) , 215-224
Partial Multivariate Errors-in-Variables Model and Its Application in Settlement Monitoring
Aiming at the problem that the coefficient matrix of multivariate errors-in-variables (MEIV) model contains constant columns, the MEIV model is extended to Partial multivariate errors-in-variables (P-MEIV), and the new algorithm of P-MEIV model is proposed based on the principle of Partial errors-in-variables (PEIV) model and indirect adjustment. The algorithm is simple and easy to implement. An example of coordinate transformation is used for verifying, and the results are compared with the existing MEIV model algorithm, which shows the effectiveness of the proposed algorithm. Finally, the P-MEIV algorithm is applied to the multi-point grey model (MGM(1,N)) of settlement monitoring. The results show that the P-MEIV model proposed in this paper can better consider the influence of monitoring point errors, and the estimated results are in good agreement with the actual situation.
- 13 June 2024
- Published by Taylor & Francis in Journal of Spatial Science
- Vol. 70 (1) , 103-120
Estimation and analysis of GNSS specific-observation biases using constrained rank-deficient virtual observation algorithm
- 13 June 2024
- Published by Taylor & Francis in Journal of Spatial Science
- Vol. 70 (1) , 103-120
Estimation and analysis of GNSS specific-observation biases using constrained rank-deficient virtual observation algorithm
- 3 May 2024
- Published by Elsevier in Advances In Space Research
- Vol. 74 (3) , 1471-1481
Analyzing the spatial variation characteristics of grid TEC using long-term GIM data
- 3 May 2024
- Published by Elsevier in Advances In Space Research
- Vol. 74 (3) , 1471-1481
Analyzing the spatial variation characteristics of grid TEC using long-term GIM data
- Open Access
- 20 February 2024
- Published by Copernicus GmbH in Annales Geophysicae
- Vol. 42 (1) , 45-53
Characteristic analysis of the differences between total electron content (TEC) values in global ionosphere map (GIM) grids
Using total electron content (TEC) from a global ionosphere map (GIM) for ionospheric delay correction is a common method of eliminating ionospheric errors in satellite navigation and positioning. On this basis, the TEC of a puncture point can be obtained by GIM grid TEC interpolation. However, in terms of grid, only few studies have analyzed the TEC value size characteristics of its four grid points, that is, the TEC difference characteristics among them. In view of this, by utilizing the GIM data from high solar-activity years (2014) and low solar-activity years (2021) provided by CODE (Center for Orbit Determination in Europe), this paper proposes the grid TEC difference as a way of analyzing TEC variation characteristics within the grid, which is conducive to exploring and analyzing the variation characteristics of the ionosphere TEC in the single-station area. The value is larger in high solar-activity years and generally small in low solar-activity years, and the value of high-latitude areas is always smaller than that of low-latitude areas. Specifically, in high solar-activity years, most of the GIM grid TEC internal differences are within 4 TECu (1 TECu = 1016 electrons m−2) in high-latitude and midlatitude regions, while only 78.17 % are in low-latitude regions. In low solar-activity years, the TEC difference values within a GIM grid are mostly less than 2 TECu, and most of them in the high and middle latitudes are within 1 TECu. The main finding of this analysis is that the grid TEC differences are small for most GIM grids, especially in the midlatitudes to high latitudes of low solar years. This means that relevant extraction methods and processes can be simplified when TEC within these GIM grids is needed.
- 20 February 2024
- Published by Copernicus GmbH in Annales Geophysicae
- Vol. 42 (1) , 45-53
Open AccessCharacteristic analysis of the differences between total electron content (TEC) values in global ionosphere map (GIM) grids
Using total electron content (TEC) from a global ionosphere map (GIM) for ionospheric delay correction is a common method of eliminating ionospheric errors in satellite navigation and positioning. On this basis, the TEC of a puncture point can be obtained by GIM grid TEC interpolation. However, in terms of grid, only few studies have analyzed the TEC value size characteristics of its four grid points, that is, the TEC difference characteristics among them. In view of this, by utilizing the GIM data from high solar-activity years (2014) and low solar-activity years (2021) provided by CODE (Center for Orbit Determination in Europe), this paper proposes the grid TEC difference as a way of analyzing TEC variation characteristics within the grid, which is conducive to exploring and analyzing the variation characteristics of the ionosphere TEC in the single-station area. The value is larger in high solar-activity years and generally small in low solar-activity years, and the value of high-latitude areas is always smaller than that of low-latitude areas. Specifically, in high solar-activity years, most of the GIM grid TEC internal differences are within 4 TECu (1 TECu = 1016 electrons m−2) in high-latitude and midlatitude regions, while only 78.17 % are in low-latitude regions. In low solar-activity years, the TEC difference values within a GIM grid are mostly less than 2 TECu, and most of them in the high and middle latitudes are within 1 TECu. The main finding of this analysis is that the grid TEC differences are small for most GIM grids, especially in the midlatitudes to high latitudes of low solar years. This means that relevant extraction methods and processes can be simplified when TEC within these GIM grids is needed.
- Open Access
- 19 September 2023
- Published by MDPI AG in Remote Sensing
- Vol. 15 (18) , 4607
Ionosphere Total Electron Content Modeling and Multi-Type Differential Code Bias Estimation Using Multi-Mode and Multi-Frequency Global Navigation Satellite System Observations
With the rapid development of multi-mode and multi-frequency GNSSs (including GPS, GLONASS, BDS, Galileo, and QZSS), more observations for research on ionosphere can be provided. The Global Ionospheric Map (GIM) products are generated based on the observation of multi-mode and multi-frequency GNSSs, and comparisons with other GIMs provided by the ionosphere analysis centers are provided in this paper. Taking the CODE (Center of Orbit Determination in Europe) GIM as a reference during 30 days in January 2019, for the GIMs from JPL (Jet Puls Laboratory), UPC (Technical University of Catalonia), ESA (European Space Agency), WHU (Wuhan University), CAS (Chinese Academy of Sciences), and MMG (The multi-mode and multi-frequency GNSS observations used in this paper), the mean bias with respect to CODE products is 1.87, 1.30, −0.10, 0.01, −0.02, and −0.71 TECu, and the RMS is 2.12, 2.00, 1.33, 0.88, 0.88, and 1.30 TECu, respectively. The estimated multi-type DCB is also in good agreement with the DCB products provided by the MGEX.
- 19 September 2023
- Published by MDPI AG in Remote Sensing
- Vol. 15 (18) , 4607
Open AccessIonosphere Total Electron Content Modeling and Multi-Type Differential Code Bias Estimation Using Multi-Mode and Multi-Frequency Global Navigation Satellite System Observations
With the rapid development of multi-mode and multi-frequency GNSSs (including GPS, GLONASS, BDS, Galileo, and QZSS), more observations for research on ionosphere can be provided. The Global Ionospheric Map (GIM) products are generated based on the observation of multi-mode and multi-frequency GNSSs, and comparisons with other GIMs provided by the ionosphere analysis centers are provided in this paper. Taking the CODE (Center of Orbit Determination in Europe) GIM as a reference during 30 days in January 2019, for the GIMs from JPL (Jet Puls Laboratory), UPC (Technical University of Catalonia), ESA (European Space Agency), WHU (Wuhan University), CAS (Chinese Academy of Sciences), and MMG (The multi-mode and multi-frequency GNSS observations used in this paper), the mean bias with respect to CODE products is 1.87, 1.30, −0.10, 0.01, −0.02, and −0.71 TECu, and the RMS is 2.12, 2.00, 1.33, 0.88, 0.88, and 1.30 TECu, respectively. The estimated multi-type DCB is also in good agreement with the DCB products provided by the MGEX.