Answers are grounded in this paper and cite their sources.
As a global dietary staple, rice production faces escalating threats from the rice leaf folder (RLF), which has emerged as a significant biotic threat across Asia. This growing pressure has intensified the demand for scalable monitoring in the field. Traditional monitoring relies heavily on manual scouting, which is labor-intensive and difficult to scale. More advanced monitoring approaches, however, often require hyperspectral imaging systems that are too costly for routine field deployment. To overcome these limitations, this study developed RLFgoX. This early warning system uses low-cost RGB imagery acquired by UAVs to detect RLF infestation during the rice tillering stage. The proposed method adopts a rice cluster-based detection strategy, where each rice cluster is treated as the fundamental analytical unit. First, the Green-Red Vegetation Index (GRVI) was applied to reduce background interference. A comprehensive feature generation step was then performed to extract five feature types describing RLF feeding-induced mechanical damage and environmental conditions. These features were used as input datasets for machine learning classification of RLF-Infested and Non-Infested clusters. By integrating a random forest classifier with data down-sampling and a fuzzy distance-based ensemble method, the RLFgoX model achieved a Matthews correlation coefficient of 0.454, with a sensitivity of 0.756 and specificity of 0.733. Additional field validation showed that the prediction outputs were generally consistent with expert annotations and field observations, supporting the practical applicability of the proposed system for early damage detection under field conditions. By converting these results into high-resolution spatial risk heat maps, RLFgoX facilitates site-specific pest management and provides a scalable, cost-effective framework for sustainable precision agriculture.
No takes yet. Share an insight, caveat, or question.
Chen et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: