A Clustering–Connection Algorithm for Coarse Root System Architecture Reconstruction Based on Ground-Penetrating Radar

Abstract

Root system architecture (RSA), a critical attribute of plant roots, necessitates in situ reconstruction to advance the understanding of the subsurface plant root system. Ground Penetrating Radar (GPR), a non-invasive geophysical technique for in situ detection, has demonstrated success in plant RSA reconstruction. However, existing GPR-based methods have limitations, including their applicability to a specific survey line arrangement, reliance on root attribute information, numerous parameter settings, and a focus on incomplete root systems. To address these issues, a new clustering–connection (CC) method is proposed, which considers the root extension direction and growth characteristics for RSA reconstruction. Experimental results show that the CC method achieves accuracy rates of 93.38% and 88.17% for circular and grid survey line arrangements in simulated data, with deviation rates of 3.23% and 9.17% for root lengths. The method also delivered effective results with measured data. This study overcomes the limitations of survey lines and numerous parameters, enabling effective RSA reconstruction. It provides a methodological foundation and reference data for using GPR in urban tree root monitoring by estimating ecological parameters in the forest subsurface and analyzing root distribution patterns in deep-rooted and shallow-rooted plants.

Plant phenotyping relevance

GPRを用いて植物の根系構造を再構成する新規クラスタリング・接続手法を開発し、シミュレーションおよび実測データで精度検証しているため、根系形態の取得・推定が中心的です。

Code and data availability

The supplied blocks describe GPR-based root system architecture reconstruction with simulated and field-measured data, but contain no data availability statement, no public repository deposit, and no author code/model release. matGPR is a cited third-party tool, not a paper-specific asset. No qualifying public asset is

No evidence-backed public reproduction asset is currently recorded.