Comparative study on leaf disease identification using Yolo v4 and Yolo v7 algorithm

Agriculture is the primary occupation of almost all countries which provides food to the world population. The population explosion and growing demands for food necessitates the farmers to increase the food production to cater the needs. On the other hand farming is not viewed as an profitable profession, as the farmers suffers heavy loss due to pests and diseases that affects the quality and quantity of the farm produces. Hence, prediction of plant diseases in early stage using contemporary technologies will help the farmers to take well informed early decisions. This work uses and compares the results of two important Computer vision algorithms namely YOLOv4 and YOLOv7 in classifying the leaf diseases from the leaf images of variety of plant species. The models are trained with individual leaf images shot under different ambience, which imparts robustness and versatility to the models. Both the models effectively annotate and predict the leaf disease with good confidence score for each class. The other classification metrics like Precision, F1- score, Mean Average Precision, and recall also shows competitive results. However, YOLOv7 exhibits comparatively better performance as it dynamically learns the class labels through its soft labelling mechanism. Also, the work can be extended in future to predict the extent of damage with recommendation strategies.

Text Comparative study on leaf disease identification using Yolo v4 and Yolo v7 algorithm.pdf - Published Version Restricted to Repository staff only Download (1MB) Official URL or Download Paper: https://agbioforum.org/menuscript/index.php/agb/ar...

Actions (login required)

View Item