Leijun Xu
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
Leijun Xu has a total of 65 co-authors. Here are the top co-authors by number of shared publications:
- Xue Bai
School of Electrical and Information Engineering, Jiangsu University, Zhenjiang 212013, China
11 shared publications Xue Bai
9 shared publications- Jianfeng Chen
School of Electrical and Information Engineering, Jiangsu University, Zhenjiang 212013, People’s Republic of China
7 shared publications Jianfeng Chen
School of Electrical and Information Engineering, Jiangsu University, Zhenjiang 212013, China
6 shared publications- Xinke Zhao
School of Electrical and Information Engineering, Jiangsu University, Zhenjiang 212013, People’s Republic of China
5 shared publications - Xue Bai
School of Electrical and Information Engineering, Jiangsu University, Zhenjiang 212013, People’s Republic of China
5 shared publications
Latest Publications
- Open Access
- 13 October 2026
- Published by Elsevier in SSRN Electronic Journal
Spectral Structure Synergistic Optimization Improves Near-Infrared Prediction of Fatty Acid Value in Stored Grains
To improve the predictive accuracy and variable-selection stability of near-infrared (NIR) spectroscopy for determining fatty acid value (FAV) in stored grains, this study investigated stored rice and wheat flour using a portable NIR spectrometer covering the wavelength range of 899.22–1724.71 nm, with FAV determined by anhydrous ethanol extraction followed by potassium hydroxide titration as the reference method. A physics-informed spectral structure synergistic optimization method (SSSO2) was introduced to organize wavelength variables into adaptive, non-uniform spectral structures. Informative variable subsets were subsequently identified through structure-combination searching and intra-structure pruning, and the performance of SSSO2 was compared with that of competitive adaptive reweighted sampling (CARS), interval partial least squares (iPLS), and interval variable iterative space shrinkage approach (iVISSA). In the final external validation, SSSO2 achieved prediction-set coefficients of determination of 0.942 and 0.918 for rice and wheat flour, respectively, with corresponding root mean square errors of prediction of 0.737 and 1.440 mg KOH/100 g, outperforming the comparative methods. Across repeated runs, SSSO2 selected an identical variable subset for wheat flour and also showed high variable-selection consistency for rice. By preserving informative wavelength regions with coherent spectral-structural characteristics, SSSO2 achieved a favorable balance among predictive performance, selection stability, and variable parsimony. These results demonstrate its potential as a robust variable-selection strategy for the rapid and nondestructive determination of FAV in stored grains.
- 13 October 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessSpectral Structure Synergistic Optimization Improves Near-Infrared Prediction of Fatty Acid Value in Stored Grains
To improve the predictive accuracy and variable-selection stability of near-infrared (NIR) spectroscopy for determining fatty acid value (FAV) in stored grains, this study investigated stored rice and wheat flour using a portable NIR spectrometer covering the wavelength range of 899.22–1724.71 nm, with FAV determined by anhydrous ethanol extraction followed by potassium hydroxide titration as the reference method. A physics-informed spectral structure synergistic optimization method (SSSO2) was introduced to organize wavelength variables into adaptive, non-uniform spectral structures. Informative variable subsets were subsequently identified through structure-combination searching and intra-structure pruning, and the performance of SSSO2 was compared with that of competitive adaptive reweighted sampling (CARS), interval partial least squares (iPLS), and interval variable iterative space shrinkage approach (iVISSA). In the final external validation, SSSO2 achieved prediction-set coefficients of determination of 0.942 and 0.918 for rice and wheat flour, respectively, with corresponding root mean square errors of prediction of 0.737 and 1.440 mg KOH/100 g, outperforming the comparative methods. Across repeated runs, SSSO2 selected an identical variable subset for wheat flour and also showed high variable-selection consistency for rice. By preserving informative wavelength regions with coherent spectral-structural characteristics, SSSO2 achieved a favorable balance among predictive performance, selection stability, and variable parsimony. These results demonstrate its potential as a robust variable-selection strategy for the rapid and nondestructive determination of FAV in stored grains.
- 22 September 2026
- Published by Elsevier in Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy
- Vol. 366, 128822
Fingerprint spectrum matching-empowered terahertz metamaterial sensor for the ultrasensitive detection of sorbic acid
- 22 September 2026
- Published by Elsevier in Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy
- Vol. 366, 128822
Fingerprint spectrum matching-empowered terahertz metamaterial sensor for the ultrasensitive detection of sorbic acid
- Open Access
- 26 August 2026
- Published by MDPI AG in Agriculture
- Vol. 16 (17) , 1826
A Comprehensive Review of Deep Learning in Agricultural Visual Perception: Progress, Bottlenecks, and Emerging Trends
This paper provides a comprehensive review of deep learning in agricultural visual perception. Climate change and land degradation demand a shift from experience-driven to data-driven intelligent agriculture. Traditional manual inspections remain subjective and unscalable. Computer vision offers a non-invasive solution for precision crop and livestock management, while challenges also exist. Its application in actual agricultural scenarios faces specific challenges such as severe occlusion, drastic changes in lighting, and non-rigid deformation of biological targets. To systematically summarize how these perception bottlenecks are being resolved, this review explores deep learning architectures featuring spatial extraction and spatio-temporal modeling. This comprehensive review first constructs a progressive analytical framework from low-level data augmentation and static spatial cognition to high-level dynamic spatio-temporal reasoning, and then systematically classifies existing literature. In-depth analysis reveals that these three core challenges can be computationally resolved using three unified technologies. Domain drift and label scarcity, rather than baseline accuracy, are the main obstacles to practical deployment. We further identified four priority research directions, i.e., architectural efficiency, data-level annotation efficiency, deployment-level privacy and simulation, and trust-level security, to guide future research.
- 26 August 2026
- Published by MDPI AG in Agriculture
- Vol. 16 (17) , 1826
Open AccessA Comprehensive Review of Deep Learning in Agricultural Visual Perception: Progress, Bottlenecks, and Emerging Trends
This paper provides a comprehensive review of deep learning in agricultural visual perception. Climate change and land degradation demand a shift from experience-driven to data-driven intelligent agriculture. Traditional manual inspections remain subjective and unscalable. Computer vision offers a non-invasive solution for precision crop and livestock management, while challenges also exist. Its application in actual agricultural scenarios faces specific challenges such as severe occlusion, drastic changes in lighting, and non-rigid deformation of biological targets. To systematically summarize how these perception bottlenecks are being resolved, this review explores deep learning architectures featuring spatial extraction and spatio-temporal modeling. This comprehensive review first constructs a progressive analytical framework from low-level data augmentation and static spatial cognition to high-level dynamic spatio-temporal reasoning, and then systematically classifies existing literature. In-depth analysis reveals that these three core challenges can be computationally resolved using three unified technologies. Domain drift and label scarcity, rather than baseline accuracy, are the main obstacles to practical deployment. We further identified four priority research directions, i.e., architectural efficiency, data-level annotation efficiency, deployment-level privacy and simulation, and trust-level security, to guide future research.
- 4 August 2026
- Published by Elsevier in Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy
- Vol. 364 (Pt 1) , 128548
Homologous heterogeneity: comparative study of derivative data modes of terahertz time-domain spectroscopy in olive oil authenticity analysis
- 4 August 2026
- Published by Elsevier in Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy
- Vol. 364 (Pt 1) , 128548
Homologous heterogeneity: comparative study of derivative data modes of terahertz time-domain spectroscopy in olive oil authenticity analysis
- 28 July 2026
- Published by Springer Nature in Food Analytical Methods
- Vol. 19 (10) , 293
Rapid and Non-contact Classification of Edible Oils Using Terahertz Time-Domain Spectroscopy Combined with Pattern Recognition Techniques
- 28 July 2026
- Published by Springer Nature in Food Analytical Methods
- Vol. 19 (10) , 293
Rapid and Non-contact Classification of Edible Oils Using Terahertz Time-Domain Spectroscopy Combined with Pattern Recognition Techniques
- 10 June 2026
- Published by Elsevier in Microelectronics Journal
- Vol. 175
Design of a 915-MHz high-efficiency wide-dynamic-range RF energy harvesting circuit for wireless sensor networks
- 10 June 2026
- Published by Elsevier in Microelectronics Journal
- Vol. 175
Design of a 915-MHz high-efficiency wide-dynamic-range RF energy harvesting circuit for wireless sensor networks
- 16 March 2026
- Published by IOP Publishing in Physica Scripta
- Vol. 101 (11) , 115503
Flexible rectifying metasurfaces for efficient wireless energy harvesting
A highly efficient, ultra-thin, and flexible rectifying metasurface (RMS) operating at 5.8 GHz is presented. The unit cell utilizes a composite symmetric structure. This structure consists of a double-cross and a notched rectangular ring. This geometric symmetry ensures polarization insensitivity. Structural parameters were optimized to achieve simultaneous impedance matching. The metasurface matches the impedance of free space ( Z 0 ) and provides a conjugate match to the diodes. This eliminates the need for external matching networks, thereby significantly enhancing system integration. Furthermore, the high-Q chip inductors within the interconnect units serve both as direct current (DC) pathways and, in conjunction with the unit’s inherent parasitic capacitance, form radio frequency (RF) filters. This coplanar integration obviates the need for a separate power combining network. A 6×6 array (96 mm × 96 mm) was fabricated on a 0.6 mm thick polyimide substrate. Experimental measurements demonstrate a total efficiency exceeding 67% across all polarization angles. The efficiency remains above 64.2% within an incident angle range of 0° to 30°. Additionally, the system maintains an efficiency of 66.74% when bent to a curvature radius of 180 mm. These results demonstrate excellent conformability. These attributes make the proposed RMS a highly efficient and compact solution for wireless power delivery, particularly for Internet of Things (IoT) and wearable applications.
- 16 March 2026
- Published by IOP Publishing in Physica Scripta
- Vol. 101 (11) , 115503
Flexible rectifying metasurfaces for efficient wireless energy harvesting
A highly efficient, ultra-thin, and flexible rectifying metasurface (RMS) operating at 5.8 GHz is presented. The unit cell utilizes a composite symmetric structure. This structure consists of a double-cross and a notched rectangular ring. This geometric symmetry ensures polarization insensitivity. Structural parameters were optimized to achieve simultaneous impedance matching. The metasurface matches the impedance of free space ( Z 0 ) and provides a conjugate match to the diodes. This eliminates the need for external matching networks, thereby significantly enhancing system integration. Furthermore, the high-Q chip inductors within the interconnect units serve both as direct current (DC) pathways and, in conjunction with the unit’s inherent parasitic capacitance, form radio frequency (RF) filters. This coplanar integration obviates the need for a separate power combining network. A 6×6 array (96 mm × 96 mm) was fabricated on a 0.6 mm thick polyimide substrate. Experimental measurements demonstrate a total efficiency exceeding 67% across all polarization angles. The efficiency remains above 64.2% within an incident angle range of 0° to 30°. Additionally, the system maintains an efficiency of 66.74% when bent to a curvature radius of 180 mm. These results demonstrate excellent conformability. These attributes make the proposed RMS a highly efficient and compact solution for wireless power delivery, particularly for Internet of Things (IoT) and wearable applications.