Synthetic Data Augmentation and Deep Learning for Real-Time Weed Detection in Agricultural Fields

M I Nuha Marzuqha; Tshewang Rigzin; Sonam Tharchen

doi:10.36948/ijfmr.2024.v06i06.31555

Synthetic Data Augmentation and Deep Learning for Real-Time Weed Detection in Agricultural Fields

Author(s)	M I Nuha Marzuqha, Tshewang Rigzin, Sonam Tharchen
Country	India
Abstract	The ability to identify weeds in agricultural fields is critical to increasing productivity of crops as well as minimizing the use of herbicides. This paper presents a combined approach focusing on crop-weed discrimination employing the object detection capabilities of YOLOv8, the image classification power of VGG16 and, Grad-CAM, for the purpose of understanding the concerns of the model predictions. With a custom dataset of images consisting of mixed and clean images of potato and carrot crops, enhanced data limitations are achieved by training a CycleGAN model to synthesize clean carrot images from carrot-weed composite images, thus augmenting the dataset and allowing for better model performance in different settings. The pipeline begins with a vision object detection network called YOLOv8, which is used to detect the crops and the weeds in the image by drawing bounding boxes around the areas of interest. VGG16 then takes it a step further by classifying the regions, specifically differentiating between crops and weeds even more accurately. Classification outcomes are further enhanced by Grad-CAM which helps to visualize and elucidate the classifications giving an understanding of an area of interest in the models’ prediction. when assessed across the various metrics, the combined strategy improves the precision and recall measures over the single model systems. This two-model modular design avails a suitable approach for weed detection in the fields in a real-time situation and can be extended for crop monitoring and precision agriculture purposes.
Keywords	Precision Agriculture, Weed Detection, YOLOv8, VGG16, CycleGAN, Grad-CAM, Crop Management, Object Detection, Image Classification, Machine Learning, Synthetic Data Augmentation.
Field	Computer > Artificial Intelligence / Simulation / Virtual Reality
Published In	Volume 6, Issue 6, November-December 2024
Published On	2024-11-25
Cite This	Synthetic Data Augmentation and Deep Learning for Real-Time Weed Detection in Agricultural Fields - M I Nuha Marzuqha, Tshewang Rigzin, Sonam Tharchen - IJFMR Volume 6, Issue 6, November-December 2024. DOI 10.36948/ijfmr.2024.v06i06.31555
DOI	https://doi.org/10.36948/ijfmr.2024.v06i06.31555
Short DOI	https://doi.org/g8r8hb

View / Download PDF File

E-ISSN 2582-2160

doi

CrossRef DOI is assigned to each research paper published in our journal.

IJFMR DOI prefix is
10.36948/ijfmr

Downloads

Research Paper Format Copyright Permission Form and Undertaking Form Cover Page Vol 6 Isu 6 Cover Page Vol 6 Isu 5 Cover Page Vol 6 Isu 4

All research papers published on this website are licensed under Creative Commons Attribution-ShareAlike 4.0 International License, and all rights belong to their respective authors/researchers.

CC-BY-SA

About IJFMR Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us		+91-9898-5948-55	editor@ijfmr.com

International Journal For Multidisciplinary Research

A Widely Indexed Open Access Peer Reviewed Multidisciplinary Bi-monthly Scholarly International Journal

Synthetic Data Augmentation and Deep Learning for Real-Time Weed Detection in Agricultural Fields

Share this