An Improved Analysis of Weed Prediction in Precision Agriculture using Deep Learning

The proliferation of weeds is a formidable obstacle for farmers in agriculture, resulting in diminished crop productivity and elevated expenses associated with their management. To tackle this problem, precision agricultural methods have been created to detect and control weeds promptly and effectiv...

Full description

Saved in:
Bibliographic Details
Published in2024 International Conference on Communication, Computer Sciences and Engineering (IC3SE) pp. 936 - 941
Main Authors Pal, Atul Kumar, Patro, B.D.K, Chaudhary, Preeti, Richa, Kumar, Anuj, Garg, Navin
Format Conference Proceeding
LanguageEnglish
Published IEEE 09.05.2024
Subjects
Accuracy
Agriculture
Biological system modeling
CNN
Crops
Deep learning
Drones
Precision agriculture
Productivity
Training
Weed
Online AccessGet full text

Cover

More Information
Summary:The proliferation of weeds is a formidable obstacle for farmers in agriculture, resulting in diminished crop productivity and elevated expenses associated with their management. To tackle this problem, precision agricultural methods have been created to detect and control weeds promptly and effectively precisely. Deep learning is utilized to train a convolutional neural network (CNN) to identify patterns and characteristics in photographs of crop fields, enabling it to differentiate between weeds and crops. At first, drones or other airborne photography technologies will be used to gather high-resolution pictures of the crop fields. The photos are subsequently subjected to pre-processing techniques to eliminate noise and improve the characteristics of the crops and weeds. Afterward, the images are inputted into the model, which has several layers of artificial neurons that acquire the ability to extract characteristics and categorize the input images. During the training phase, the CNN utilizes a substantial dataset of labelled photos to adjust its parameters and enhance its precision in differentiating weeds and crops. After the model has been trained, it may be utilized to accurately forecast the existence of weeds in novel photographs of agricultural areas. Farmers may effectively identify and focus on specific regions for weed management by using this technology, decreasing the number of herbicides required and mitigating any ecological damage.
DOI:10.1109/IC3SE62002.2024.10593570