A Survey on 3D Reconstruction Techniques in Plant Phenotyping: From Classical Methods to Neural Radiance Fields (NeRF), 3D Gaussian Splatting (3DGS), and Beyond

2025-04-30

AI Summary

Accurate 3D reconstruction of plants remains challenging in real-world agricultural settings due to sparse, noisy, and uncontrolled outdoor imaging conditions. Method: This study systematically evaluates classical multi-view stereo (MVS), neural radiance fields (NeRF), and the emerging 3D Gaussian Splatting (3DGS) for plant phenotyping—introducing 3DGS into the plant reconstruction taxonomy for the first time. We conduct a rigorous comparative analysis across data sparsity, noise robustness, outdoor adaptability, and computational efficiency, and propose an agriculture-oriented method selection framework grounded in empirical performance boundaries. Contribution/Results: We establish the first comprehensive methodology benchmark for plant 3D reconstruction, release an open-source evaluation benchmark—including a multimodal dataset, standardized evaluation protocols, and practical guidelines (publicly available on GitHub)—and demonstrate how 3DGS enables scalable, automated phenotyping. This work advances both methodological rigor in agricultural computer vision and the paradigmatic adoption of 3DGS for field-deployable plant sensing.

Technology Category

• Computer Vision: 3D Computer Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Plant phenotyping plays a pivotal role in understanding plant traits and their interactions with the environment, making it crucial for advancing precision agriculture and crop improvement. 3D reconstruction technologies have emerged as powerful tools for capturing detailed plant morphology and structure, offering significant potential for accurate and automated phenotyping. This paper provides a comprehensive review of the 3D reconstruction techniques for plant phenotyping, covering classical reconstruction methods, emerging Neural Radiance Fields (NeRF), and the novel 3D Gaussian Splatting (3DGS) approach. Classical methods, which often rely on high-resolution sensors, are widely adopted due to their simplicity and flexibility in representing plant structures. However, they face challenges such as data density, noise, and scalability. NeRF, a recent advancement, enables high-quality, photorealistic 3D reconstructions from sparse viewpoints, but its computational cost and applicability in outdoor environments remain areas of active research. The emerging 3DGS technique introduces a new paradigm in reconstructing plant structures by representing geometry through Gaussian primitives, offering potential benefits in both efficiency and scalability. We review the methodologies, applications, and performance of these approaches in plant phenotyping and discuss their respective strengths, limitations, and future prospects (https://github.com/JiajiaLi04/3D-Reconstruction-Plants). Through this review, we aim to provide insights into how these diverse 3D reconstruction techniques can be effectively leveraged for automated and high-throughput plant phenotyping, contributing to the next generation of agricultural technology.

Problem

• Review 3D reconstruction techniques for plant phenotyping
• Compare classical methods, NeRF, and 3DGS approaches
• Address challenges in accuracy, scalability, and computational cost

Innovation

• Classical methods use high-resolution sensors for flexibility
• NeRF enables photorealistic 3D from sparse viewpoints
• 3DGS represents geometry via efficient Gaussian primitives