Junguo Hu

Prof. Dr. Junguo Hu


Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
27 September 2023 in Remote Sensing
Verified
Journal Article
Open access
PDF file
JHJinjing Hou
Jinjing Hou
HZHoukui Zhou
Houkui Zhou
JHJunguo Hu
Junguo Hu
HYHuimin Yu
Huimin Yu
Hao HuHao Hu
Hao Hu
Forest tree species identification in the field of remote sensing has become an important research topic. Currently, few research methods combine global and local features, making it challenging to accurately handle the similarity between different categories. Moreover, using a single deep layer for feature extraction overlooks the unique feature information at intermediate levels. This paper proposes a remote sensing image forest tree species classification method based on the Multi-Scale Convolution and Multi-Level Fusion Network (MCMFN) architecture. In the MCMFN network, the Shallow Multi-Scale Convolution Attention Combination (SMCAC) module replaces the original 7 × 7 convolution at the first layer of ResNet-50. This module uses multi-scale convolution to capture different receptive fields, and combines it with the attention mechanism to effectively enhance the ability of shallow features and obtain richer feature information. Additionally, to make efficient use of intermediate and deep-level feature information, the Multi-layer Selection Feature Fusion (MSFF) module is employed to improve classification accuracy. Experimental results on the Aerial forest dataset demonstrate a classification accuracy of 91.03%. The comprehensive experiments indicate the feasibility and effectiveness of the proposed MCMFN network.

2 February 2023 in Sustainability
Verified
Journal Article
Open access
PDF file
THTao He
Tao He
HZHoukui Zhou
Houkui Zhou
Caiyao XuCaiyao Xu
Caiyao Xu
JHJunguo Hu
Junguo Hu
XXXingyu Xue
Xingyu Xue
QWQun Wang
Qun Wang
Forest tree species information plays an important role in ecology and forest management, and deep learning has been used widely for remote sensing image classification in recent years. However, forest tree species classification using remote sensing images is still a difficult task. Since there is no benchmark dataset for forest tree species, a forest tree species dataset (FTSD) was built in this paper to fill the gap based on the Sentinel-2 images. The FTSD contained nine kinds of forest tree species in Qingyuan County with 8,815 images, each with a resolution of 64 × 64 pixels. The images were produced by combining forest management inventory data and Sentinel-2 images, which were acquired with less than 20% clouds from 1 April to 31 October, including the years 2017, 2018, 2019, 2020, and 2021. Then, the images were preprocessed and downloaded from Google Earth Engine (GEE). Four different band combinations were compared in the paper. Moreover, a Principal Component Analysis (PCA) and Normalized Difference Vegetation Index (NDVI) were also calculated using the GEE. Deep learning algorithms including DenseNet, EfficientNet, MobileNet, ResNet, and ShuffleNet were trained and validated in the FTSD. RGB images with red, green, and blue (PC1, PC2, and NDVI) obtained the highest validation accuracy in four band combinations. ResNet obtained the highest validation accuracy in all algorithms after 500 epochs were trained in the FTSD, which reached 84.91%. As a famous and widely used remote sensing classification satellite imagery dataset, NWPU RESISC-45 was also trained and validated in the paper. ResNet achieved a high validation accuracy of 87.90% after training 100 epochs in NWPU RESISC-45. The paper shows in forest tree species classification based on remote sensing images and deep learning that (1) PCA and NDVI can be combined to improve the accuracy of classification; (2) ResNet is more suitable than other deep learning algorithms including DenseNet, EfficientNet, MobileNet, and ShuffleNet in remote sensing classification; and (3) being too shallow or deep in ResNet does not perform better in the FTSD, that is, 50 layers are better than 34 and 101 layers.

9 June 2024 in Scientific Reports
Verified
Journal Article
Open access
PDF file
JLJiangnan Liu
Jiangnan Liu
JHJunguo Hu
Junguo Hu
HLHaoqi Liu
Haoqi Liu
KHKanglai Han
Kanglai Han
Soil respiration (Rs) represents the greatest carbon dioxide flux from terrestrial ecosystems to the atmosphere. However, its environmental drivers are not fully understood, and there are still significant uncertainties in soil respiration model estimates. This study aimed to estimate the spatial distribution pattern and driving mechanism of global soil respiration by constructing a machine learning model method based on ecological big data. First, we constructed ecological big data containing five categories of 27-dimensional environmental factors. We then used four typical machine learning methods to develop the performance of machine learning models under four training strategies and explored the relationship between soil respiration and environmental factors. Finally, we used the RF machine learning algorithm to estimate the global Rs spatial distribution pattern in 2021, driven by multiple dimensions of environmental factors, and derived the annual soil respiration values. The results showed that RF performed better under the four training strategies, with a coefficient of determination R2 = 0.78216, root mean squared error (RMSE) = 285.8964 gCm−2y−1, and mean absolute error (MAE) = 180.4186 gCm−2y−1, which was more suitable for the estimation of large-scale soil respiration. In terms of the importance of environmental factors, unlike previous studies, we found that the influence of geographical location was greater than that of MAP. Another new finding was that enhanced vegetation index 2 (EVI2) had a higher contribution to soil respiration estimates than the enhanced vegetation index (EVI) and normalized vegetation index (NDVI). Our results confirm the potential of utilizing ecological big data for spatially large-scale Rs estimations. Ecological big data and machine learning algorithms can be considered to improve the spatial distribution patterns and driver analysis of Rs.

14 June 2025 in Sensors
Verified
Journal Article
Open access
PDF file
HZHoukui Zhou
Houkui Zhou
CCCHANG CHENG
CHANG CHENG
ZXZhongyi Xia
Zhongyi Xia
QDQifeng Ding
Qifeng Ding
QLQinqin Liao
Qinqin Liao
THTao He
Tao He
With the rise of deep learning technology, intelligent garbage detection provides a new idea for garbage classification management. However, due to the interference of complex environments, coupled with the influence of the irregular features of garbage, garbage detection in complex scenarios still faces significant challenges. Moreover, some of the existing research suffer from shortcomings in either their precision or real-time performance, particularly when applied to complex garbage detection scenarios. Therefore, this paper proposes a model based on YOLOv8, namely HGCS-Det, for detecting garbage in complex scenarios. The HGCS-Det model is designed as follows: Firstly, the normalization attention module is introduced to calibrate the model’s attention to targets and to suppress the environmental noise interference information. Additionally, to weigh the attention-feature contributions, an Attention Feature Fusion module is employed to complement the attention weights of each channel. Subsequently, an Instance Boundary Reinforcement module is established to capture the fine-grained features of garbage by combining strong gradient information with semantic information. Finally, the Slide Loss function is applied to dynamically weight hard samples arising from the complex detection environments to improve the recognition accuracy of hard samples. With only a slight increase in parameters (3.02M), HGCS-Det achieves a 93.6% mean average precision (mAP) and 86 FPS on the public HGI30 dataset, which is a 3.33% higher mAP value than from YOLOv12, and outperforms the state-of-the-art (SOTA) methods in both efficiency and applicability. Notably, HGCS-Det maintains a lightweight architecture while enhancing the detection accuracy, enabling real-time performance even in resource-constrained environments. These characteristics significantly improve its practical applicability, making the model well suited for deployment in embedded devices and real-world garbage classification systems. This method can serve as a valuable technical reference for the engineering application of garbage classification.

7 October 2022 in Sensors
Verified
Journal Article
Open access
PDF file
XWxuecheng wu
xuecheng wu
HZHoukui Zhou
Houkui Zhou
HYHuimin Yu
Huimin Yu
Hao HuHao Hu
Hao Hu
GZGuangqun Zhang
Guangqun Zhang
THTao He
Tao He
An algorithm for a sharpness evaluation of microscopic images based on non-subsampled shearlet wave transform (NSST) and variance is proposed in the present study for the purpose of improving the noise immunity and accuracy of a microscope’s image autofocus. First, images are decomposed with the NSST algorithm; then, the decomposed sub-band images are subjected to variance to obtain the energy of the sub-band coefficients; and finally, the evaluation value is obtained from the ratio of the energy of the high- and low-frequency sub-band coefficients. The experimental results show that the proposed algorithm delivers better noise immunity performance than other methods reviewed by this study while maintaining high sensitivity.

21 August 2023 in Sustainability
Verified
Journal Article
Open access
PDF file
QSQisong Sun
Qisong Sun
JHJunguo Hu
Junguo Hu
JJJunjie Jiang
Junjie Jiang
KGKechen Gu
Kechen Gu
CZChao Zhu
Chao Zhu
WYWenjie Yin
Wenjie Yin
Purpose: The soil air diffusion coefficient (Ds) is particularly important for the study of soil gas diffusion movement, and there are still many uncertainties in the widely used methods; as such, a method was designed to in situ measure the soil gas diffusion coefficient. Methods: Four different soil media were selected and studied by means of a designed in situ measurement of soil gas diffusion coefficients, and these were compared and analyzed with the predictions of several commonly used prediction models. In addition, they were combined with gas transport models to validate the results of the empirical models that were obtained by the in situ measurements. Results: The results of the data indicate that increasing the volumetric soil moisture content decreases the soil gas diffusion coefficient, with changes in the soil gas diffusion coefficient for the small-grained quartz sand medium being similar to those predicted by the Buckingham model. The soil gas diffusion coefficients for the large-grained quartz sand were similar to the Millington and Quirk model predictions at low humidity; for increased humidity, it was instead similar to the Buckingham model predictions. The soil gas diffusion coefficients of the two active media were closer to those of the SWLR model with high Cm. In addition, the R2 of the measured data was verified, by empirical modeling, to be greater than 0.54, and inversion experiments were conducted to verify that the results were consistent with those of the SWLR model. Conclusion: When measuring relative diffusion coefficients in the field, we recommend the in situ measurement method, which is more reflective of the actual situation in natural environments and provides more accurate data support for soil carbon flux studies.