Identification and Detection of Leaf Miner, Pests Infestation in Cucurbitaceae Family in Real Time Infield Scenarios using YOLOv5s Object Detection Model

This study contributes to technological developments in precision agriculture by developing a YOLOv5sbased model with a high F1 score and mAP, highlighting its potential for real-world use in wax gourd plants. In precision agriculture, early disease identification is crucial for effective harvesting. The present research focuses on the creation and evaluation of a YOLOv5s-based model for detecting pests and leaf miner infections in wax gourd plants in real time infield conditions at various stages of plant growth. Using a dataset of over 2000 samples collected in a variety of in-field settings, the model was trained for 200 epochs with a low learning rate (0.00001) and a batch size of 32. The achieved F1 score of 0.96 demonstrates the model's remarkable combination of precision and recall, ensuring precise disease detection. Furthermore, a mean Average Precision (mAP) greater than 0.45 demonstrates the model's proficiency in precise leaf localization and overall detection ability. The model's ability to detect numerous occurrences of disease in different leaves simultaneously is a crucial strength, reflecting the complexities common in real-world agricultural scenarios. The dataset's quantity and diversity help the model be robust and generalizable across diverse infection types and growth stages.