Fig 1.
Sample images from the dataset for insect pest detection.
Fig 2.
Augmented insect pest images using geometric transformations.
Fig 3.
Image annotation process for training data using Makesense platform.
Fig 4.
Overview of the YOLOv10 architecture, Highlighting Backbone, Neck, and Head components.
Fig 5.
Overview of the improved YOLOv10 architecture with C3 and ConvTranspose2d enhancements.
Fig 6.
Comparison of feature extraction in different models:
(a) C3 block, (b) C2fCIB block, and (c) C2f block.
Fig 7.
Functional overview of the IoT-based smart insect trap.
Fig 8.
Workflow of the proposed insect trap system.
Fig 9.
Workflow of the proposed insect trap system.
Fig 10.
YOLOv5-m Results for Codling Moth Detection.
Fig 11.
YOLOv6-m Results for Codling Moth Detection.
Fig 12.
YOLOv8-m Results for Codling Moth Detection.
Fig 13.
YOLOv9-m Results for Codling Moth Detection.
Fig 14.
YOLOv10-m Results for Codling Moth Detection.
Fig 15.
YOLOv11-m Results for Codling Moth Detection.
Fig 16.
YOLOv12-m Results for Codling Moth Detection.
Fig 17.
Improved YOLOv10-m results for Codling Moth detection.
Table 1.
Performance metrics of YOLO models for Codling Moth detection.
Table 2.
Confidence and parameter metrics of YOLO models for Codling Moth detection.
Fig 18.
Illustrative example of a real-time pest monitoring dashboard accessed via a web interface.
The figure was created by the authors for visualization purposes and is conceptually similar to IoT-based monitoring dashboards. It does not reproduce or adapt any proprietary interface elements and is provided for illustrative purposes only under the Creative Commons Attribution (CC BY 4.0) license. Base map and data from OpenStreetMap and OpenStreetMap foundation.
Fig 19.
Illustrative example of a real-time pest monitoring dashboard accessed via a mobile application.
This figure was created by the authors for visualization purposes and is conceptually similar to dashboards used in IoT-based monitoring platforms. It does not reproduce, adapt, or include any proprietary interface elements from third-party platforms. The figure is provided for illustrative purposes only and is published under the Creative Commons Attribution (CC BY 4.0) license. Base map and data from OpenStreetMap and OpenStreetMap foundation.
Fig 20.
Email alert generated when pest count exceeds the configured threshold.
Fig 21.
Proposed Pesticide Recommendation System for Apple Leaf Disease Management.