Related Experiment Video
Updated: Aug 7, 2026
11:37
RGB and Spectral Root Imaging for Plant Phenotyping and Physiological Research: Experimental Setup and Imaging Protocols
Published on: August 8, 2017
Low-cost monocular RGB-based 3D structural mapping for horticultural plants via semantic scene completion
Ruijun Jing1,2, Rui An3, Zhuoxing Li3
1School of Instrument and Electronics, North University of China, Taiyuan, China.
Frontiers in Plant Science
|August 6, 2026
Summary
This study introduces a monocular 3D structural mapping framework for precision agriculture, enabling detailed crop analysis from single RGB images. The method provides accurate canopy height and volume data for enhanced crop monitoring and robotic operations.
Area of Science:
- Computer Vision
- Robotics
- Agricultural Science
Background:
- Precision agriculture requires detailed structural data (canopy height, volume) for crop monitoring and safety.
- Existing LiDAR or RGB-D sensor methods are costly and impractical for large-scale horticultural applications.
- Current 2D and SLAM-based pipelines yield insufficient semantic maps for agricultural structural analysis.
Purpose of the Study:
- To develop a monocular 3D structural mapping framework for horticultural plants using semantic scene completion.
- To enable actionable structural analysis from single RGB images, overcoming limitations of existing methods.
- To derive task-oriented structural maps, including canopy height, volume, and traversability layers.
Main Methods:
- Proposed a monocular 3D structural mapping framework via semantic scene completion.
- Introduced a Depth-Aware Decoder Module for explicit depth recovery and feature fusion.
- Designed an NCS-Guided Geometry Encoder and a Global Encoder for enhanced 3D geometric and semantic modeling.
- Utilized an RGB-only model deployed for real-world scenarios, trained on a custom horticultural 3D semantic scene dataset.
Main Results:
- Achieved high performance on a custom horticultural dataset: 82.31% occupancy IoU, 84.26% mIoU, 86.25% precision.
- Demonstrated accurate field measurements: canopy height MAE of 0.019-0.026 m and canopy volume relative errors of 8.4%-11.4%.
- Validated effectiveness on the Semantic-KITTI dataset, showcasing robustness and generalizability.
Conclusions:
- The proposed monocular 3D structural mapping framework effectively provides actionable structural insights for precision agriculture.
- The method enables cost-effective, large-scale deployment for crop monitoring and autonomous robotic operations in horticulture.
- Results confirm the framework's potential to significantly advance agricultural automation and management.
