Monitoring Bambara Groundnut Canopy State Variables at Various Growth Stages Using Low-Cost Remote Sensing Technology and Machine Learning Techniques

The 6th International Conference on Numerical Modelling in Engineering Advances in Science and Technology(2024)

引用 0|浏览8
摘要
The aim of this study was to assess the efficacy of an unmanned aerial vehicle-based remote sensing system for quantifying Bambara groundnut canopy state variables. Remotely sensed color infrared images and in-situ canopy state variables were collected during Malaysia's 2018/19 Bambara growing season at vegetative, flowering, podding, podfilling, maturity, and senescence stages. Five common vegetation indices (VIs) were derived from the images, yielding to single stage and cumulative VIs (∑VIs). The relationship between canopy state variables and single stage VIs/∑VIs was investigated using Pearson’s correlation. Linear parametric and non-linear non-parametric machine learning (ML) regressions were employed to estimate canopy state variables by using VIs/ ∑VIs as input features. The best correlation were observed at flowering stage. The ∑VIs from vegetative to senescence stage exhibited the most robust relationship with canopy state variables. CatBoostRegressor (CBR) excelled in training for all canopy state variables, however, it showed potential overfitting in testing. In contrast, Huber regression (HR) models provided consistent results in both training and testing. HR performance was comparable to that of the top-performing ML algorithms in estimation of groundnut crop variables.
更多
查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
0
您的评分 :

暂无评分

数据免责声明
页面数据均来自互联网公开来源、合作出版商和通过AI技术自动分析结果,我们不对页面数据的有效性、准确性、正确性、可靠性、完整性和及时性做出任何承诺和保证。若有疑问,可以通过电子邮件方式联系我们:report@aminer.cn