Layer I · Digitize — Physical to Digital Crop
Multi-view 3D reconstruction, UAV imaging, and computer vision turn a real crop into point clouds and quantified traits — a measurable digital twin of its current state.
Curriculum vitae · Updated July 2026
Liangchao Deng邓良超
Postdoctoral Researcher
Shenzhen Institute of China Agricultural University · Shenzhen, China
Toward high, stable, and resource-efficient crop production under climate change, my research turns satellite–UAV–ground 3D observations into interpretable crop physiological states and, by coupling structure–radiation–photosynthesis–growth processes with data assimilation, builds verifiable, predictive, and intervenable crop digital twins that provide a mechanistic basis for the coordinated design of genotype × environment × management (G×E×M). With multi-source observation and AI, I make the state of the crop system observable, its mechanisms interpretable, its future projectable, and its decision risk quantifiable.
Ph.D. in Crop Science · Shihezi University · 2026Appointment · 1 Aug 2026 – 31 Jul 2029
- architecture layers
- 4
- Ph.D. awarded
- 2026
- postdoctoral term
- 2026–29
Contact channels
Choose the right way to get in touch.
Academic correspondence, commercial projects, and administrative communication are kept separate for faster replies.
Research architecture
Four layers of crop intelligence
A four-layer architecture — Digitize, Understand, Predict, Design — turning multi-source observations into a verifiable, predictive, and designable crop. My PhD built Layers I–II; my postdoc focuses on the Layer II-to-III jump: making crop state not only observable and explainable, but projectable.
Layer II · Understand — Digital to Explainable Crop
Coupling structure–radiation–photosynthesis–growth processes with scientific AI to turn the digital crop into an interpretable, mechanistic model.
Layer III · Predict — Toward a Predictive Crop
State-transition dynamics and data assimilation project growth under environment × management scenarios, with decision risk quantified.
Layer IV · Design — Toward a Designed Crop
Inverse design and optimization over genotype × environment × management propose canopy and breeding targets — from sensing crops to designing crops.
Current appointment
Professional appointment
1 August 2026 – 31 July 2029
Postdoctoral Researcher
Shenzhen Institute of China Agricultural University, Shenzhen, China
Postdoctoral research at the Shenzhen Institute of China Agricultural University.
Academic training
Education
2021 – 2026
Ph.D. in Crop Science (Integrated Master–Ph.D.)
Shihezi University, China
- Supervisors
- Prof. Yali Zhang; Dr. Qingfeng Song; Prof. Xin-Guang Zhu
- Research focus
- Crop phenomics, UAV remote sensing, canopy photosynthesis modeling, and AI-assisted phenotyping
- Joint training
- CAS Center for Excellence in Molecular Plant Sciences (CEMPS)
- 3D crop canopy reconstruction and light-distribution simulation using SfM and 3D Gaussian Splatting.
- Fusion of RGB, multispectral, hyperspectral, and LiDAR observations.
- Deep learning for segmentation, detection, and phenotypic-parameter prediction.
2016 – 2021
B.Sc. in Information and Computational Science
Shihezi University, China
- Foundation
- Numerical analysis, computational modeling, programming, and algorithm design
- Core courses
- Computer vision, machine learning, linear algebra, optimization, graph theory, and data structures
- Graduation project
- Computational-fluid-dynamics simulation (Excellent Graduation Project)
Technical toolkit
Technical skills
Methods and tools used to move from sensing and modeling to reproducible research software.
Programming & Data
- Python · NumPy · SciPy
- PyTorch · OpenCV
- MATLAB · R
- AI product architecture
3D Vision & Point Clouds
- SfM & photogrammetry
- PCL · Open3D
- Camera calibration
- Binocular vision
Remote Sensing
- UAV imaging
- Multispectral & hyperspectral
- LiDAR processing
- Multi-source fusion
Machine Learning
- Deep learning
- Phenotypic prediction
- Statistical modeling
- AI agent workflows
Modeling & Simulation
- Ray tracing & BRDF
- Photosynthesis simulation
- Digital-twin frameworks
Research Software
- Full-stack development
- Git version control
- Algorithm modularization
Languages
- Chinese · Native
- English · Academic writing & scientific communication
Selected work
Research experience
Selected programs showing the progression from sensing and reconstruction to simulation and applied phenotyping.
AI-assisted 3D Crop Canopy Modeling
3D reconstruction, light distribution & photosynthesis
- 3D canopy reconstruction
- High-throughput farmland reconstruction using SfM, 3D Gaussian Splatting, and UAV cross-circular acquisition, reaching centimeter-level accuracy.
- Light & photosynthesis
- Canopy-scale simulation using ray tracing and BRDF-based leaf optics within a crop digital-twin framework.
- Multimodal AI
- RGB, multispectral, and LiDAR workflows for complex-scene, zero-shot plant segmentation.
- Modular research agent
- Integrated reconstruction, meshing, canopy generation, light simulation, and photosynthesis calculation into reusable modules.
High-throughput 3D & Spectral Phenotyping
Optical inversion, crop design & computer vision
- Leaf optical inversion
- Developed a BRDF-based inversion framework and optimized measurement schemes for indirect leaf-optics estimation.
- Wheat design research
- Modeled plant architecture and cultivation configurations for higher canopy efficiency, including industry-oriented collaboration.
- Phenotyping algorithms
- Built computer-vision methods for automated extraction and analysis of plant phenotypic traits.
Research outputs
Publications & software
Peer-reviewed research and reusable software supporting plant phenotyping workflows.
Peer-reviewed article2025
Leaf Bidirectional Reflectance Distribution Function (BRDF) Prediction with Phenotypic Traits in Four Species: Development of a Novel Measuring and Analyzing Framework
Plant Phenomics
Deng, L.; Yu, L. X.; Mao, L.; Wang, Y.; Guo, X.; Wang, M.; Zhang, Y.; Song, Q.; Zhu, X.-G.
Research software2025
Digital Plant Phenotyping Platform (v25.0)
Zenodo
Deng, L.
An integrated platform for plant phenotyping, data processing, and analysis. Core modules have been transferred through Shufeng Bio for applied phenotyping and intelligent-agriculture services.
