EUVIMEDEUVIMED entdecken
Crossref

Colour‐Based Soil Organic Carbon Assessment in Croplands Using In Situ RGB Imaging: A Lab‐in‐the‐Field Approach

Qi Sun, Jiwen Tian, Lulu Qi, Jun Zhang, Pu Shi

European Journal of Soil Science · 2026-09 · 77 (5)

DOI: 10.1111/ejss.70436 ↗

Kurz gefasst

Fachartikel in der Zeitschrift European Journal of Soil Science, online erschienen am 12.10.2026. Verfasst von Qi Sun und 4 weiteren Autor:innen. Die Arbeit nennt 63 Literaturangaben. Gefördert von: State Key Laboratory of Soil Erosion and Dryland Farming on the Loess Plateau, Chinese Academy of Sciences.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Advances in soil colour imaging have positioned colour‐based soil organic carbon (SOC) predictive modelling as a promising tool for direct field assessment. However, previous studies remain confined to controlled laboratory settings, lacking consideration for field‐related challenges such as variable lighting and soil moisture. This study introduces a portable ‘lab‐in‐the‐field’ imaging apparatus that standardizes illumination during data acquisition, enabling consistent RGB imagery regardless of external light conditions. 441 topsoil samples were collected across the black soil region (67,747 km 2 ) of Heilongjiang Province, Northeast China, and soil images were captured against moist field soils as well as two laboratory dry treatments (Lab‐sieved and Lab‐bulk). A set of image colour and texture features were extracted to develop and compare SOC prediction models using four algorithms (Partial Least Squares Regression, Cubist, Random Forest and eXtreme Gradient Boosting). Results show that the custom‐designed apparatus effectively eliminated natural light interference, allowing stable soil imaging across varying weather conditions. Field‐based models consistently outperformed laboratory‐based ones, with Cubist achieving the highest accuracy ( R 2 = 0.72, RMSE = 4.37 g/kg) when both colour and texture features were used as predictors. The superior performance of field models revealed that moist soils enhanced SOC‐colour relationships by strengthening correlations with chromaticity and hue parameters, while texture features further improved predictions by incorporating microscale surface roughness and soil aggregation patterns. Finally, we developed a web‐based application tool based on the top‐performing modelling strategy, demonstrating that standardized in situ RGB imaging, combined with texture‐augmented modelling, provides a low‐cost and accessible pathway for field‐based SOC assessment.

Förderung

  • State Key Laboratory of Soil Erosion and Dryland Farming on the Loess Plateau, Chinese Academy of Sciences

Zitieren

APA

Sun, Q., Tian, J., Qi, L., Zhang, J., & Shi, P. (2026). Colour‐Based Soil Organic Carbon Assessment in Croplands Using In Situ RGB Imaging: A Lab‐in‐the‐Field Approach. European Journal of Soil Science, 77(5). https://doi.org/10.1111/ejss.70436

Vancouver

Sun Q, Tian J, Qi L, Zhang J, Shi P. Colour‐Based Soil Organic Carbon Assessment in Croplands Using In Situ RGB Imaging: A Lab‐in‐the‐Field Approach. European Journal of Soil Science. 2026;77(5). doi:10.1111/ejss.70436

BibTeX
@article{Sun2026,
  author = {Sun, Qi and Tian, Jiwen and Qi, Lulu and Zhang, Jun and Shi, Pu},
  title = {Colour‐Based Soil Organic Carbon Assessment in Croplands Using In Situ RGB Imaging: A Lab‐in‐the‐Field Approach},
  journal = {European Journal of Soil Science},
  year = {2026},
  volume = {77},
  number = {5},
  doi = {10.1111/ejss.70436}
}
Kostenlos starten →