Optimizing Object Detection Models for Precision Agriculture in Boosting Plant Disease Identification

View at publisher Summarise with VIRAT
Authors: Nasim Banu Shah, Rahul Selarka, Mayur Patil, Ashutosh Gupta, Shivam Sahu, Alok Prasanna Kumar
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Agriculture plays a crucial role in India’s economy, and safeguarding crop health is vital for sustaining productivity. Plant diseases pose one of the biggest threats to crop yield and quality, leading to significant losses. Early and accurate disease identification is essential to reduce these adverse impacts. Traditionally, farmers rely on manual visual inspection to spot disease symptoms, which is often slow, labor-intensive, and requires expert knowledge. Therefore, automated disease detection methods are highly desirable. This work presents an approach using the YOLO (You Only Look Once) object detection algorithm to identify diseases in plant leaves. YOLO splits images into grids, where each segment detects objects within its region, enabling fast and efficient real-time detection. We evaluated different YOLO versions—YOLOv5, YOLOv7, YOLOv8, and YOLOv10—using the PlantDoc dataset. Our findings reveal that while YOLOv7 demands significantly more computational resources and shows unstable precision during training, YOLOv5 delivers more consistent and reliable accuracy. The latest YOLOv8 and YOLOv10 versions enhance detection precision and efficiency, with YOLOv10 excelling in balancing detection accuracy and resource consumption.

Fields of Study & Subjects

Smart Agriculture and AI Advanced Neural Network Applications Industrial Vision Systems and Defect Detection

Publication details

Journal / source
CRC Press eBooks
Pages
446–454
Publication date
25 March 2026
Language
EN
Document type
Book chapter