Answers are grounded in this paper and cite their sources.
Aphids are one of the most harmful pests in wheat production, and timely and accurate aphid monitoring is crucial for effective control. However, existing methods mainly rely on manual investigations, resulting in low quality and efficiency of pest monitoring. The main challenges in aphid monitoring include the small target size and insufficient segmentation accuracy under complex field conditions. To address these problems, we constructed an instance segmentation dataset of aphids and leaves based on 1,412 field images and proposed a leaf-scale instance segmentation model (called EVC-YOLO-Seg) by integrating the YOLO framework with the designed EV Block and Fusion Multi-scale Convolutional Attention (FMCA) module. Results showed that EVC-YOLO-Seg achieved mAP(M), mIoU, and F1-Score values of 0.888, 0.852, and 0.906, respectively. Furthermore, seven infestation indicators (quantity, density, coverage, aggregation, cluster number, mean cluster size, and max cluster size) and a comprehensive indicator (Aphid Infestation Indicator, ADI) were developed to quantitatively assess aphid infestation severity based on the model and statistical analysis. All proposed indicators showed a high correlation with actual values, with R 2 ranging from 0.814 to 0.978. By calculating eight key indicators, 157 materials were successfully classified into four categories with significantly different levels of aphid infestation, demonstrating the effectiveness of the indicators. Finally, to enhance the utilization efficiency of the algorithm and ensure that the research outcomes benefit a broader community of scientific researchers, we integrated the above methods and developed a mobile-based automatic aphid infestation analysis application named EasyAphid. Overall, the proposed method demonstrates strong performance in aphid segmentation and quantitative infestation assessment, and shows promising potential for supporting fine-grained monitoring of aphid infestation patterns in wheat under field conditions.
No takes yet. Share an insight, caveat, or question.
Deng et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: