Mesh reconstruction methods enhance 3D crop modeling quality
Evaluating Mesh Reconstruction Methods for Crop Phenotyping
Computer Vision and Pattern Recognition
Summary
Understanding how plants grow helps farmers produce more food, but it’s hard to study crops in faraway fields. The paper compares seven ways to create 3D digital models of crops from images, which lets experts examine crops remotely. They find that three methods produce better and more accurate 3D shapes of plants. This helps specialists choose the best technology for studying crops without being onsite.
What this means in practice
- •For agriculture technology developers: Select the best mesh reconstruction pipeline to build precise 3D models for remote crop monitoring and analysis.
- •For precision farming service providers: Improve crop monitoring platforms by integrating the top-performing 3D mesh methods to deliver higher-fidelity crop models to growers remotely.$Commercial implications: Enables enhanced digital crop modeling tools that offer better insights to farmers, improving decision-making services.
Authors
Karanvir Singh, Theo Morales, Binh-Son Hua, Mukesh Saini
Abstract
Phenotyping an agricultural crop is crucial for studying its entire life cycle, as it provides vital insights to improve yield and, ultimately, food production. Doing the same for crops grown on remote sites is a challenge for the specialists who cannot be available on-site. 3D reconstruction techniques offer a promising solution to this problem by enabling crop digitization, allowing specialists to access the resulting 3D crop models from anywhere at any time. In this work, we evaluate recent 3D reconstruction pipelines for crop phenotyping. We focus on 7 mesh reconstruction pipelines and measure the fidelity and consistency of their outputs qualitatively and quantitatively. Our results suggest that the meshes produced by the GGGS, PGSR, and 2DGS are preferable to the other pipelines, owing to their quantitative metrics and visually pleasing outputs. The GGGS pipeline is better than the second-best pipeline (2DGS) by about 27\% on the radar chart with 5 dimensions, namely, User ratings, Chamfer distance, LPIPS, PSNR, and SSIM.