Paper Key : IRJ************056
Author: Navale Prathmesh Rajendra
Date Published: 07 Apr 2024
Abstract
Weed management is crucial for maximizing onion crop yield and quality. However, traditional methods are labor-intensive and time-consuming. Convolutional Neural Networks (CNNs) offer a promising automated solution for weed detection. This study proposes a novel CNN-based approach tailored for onion crops.The methodology involves collecting high-resolution images of onion fields, preprocessing them to enhance contrast and eliminate noise, and training a CNN model using a large dataset of annotated images. Transfer learning and data augmentation techniques are employed to improve model performance and generalization. The proposed CNN model is evaluated on real-world onion fields, demonstrating its effectiveness in accurately detecting weeds while minimizing false positives. Comparative analysis against existing methods highlights its superior accuracy and efficiency. This study contributes to precision agriculture by offering a robust and automated solution for weed detection in onion crops. By reducing reliance on manual labor, the proposed approach promotes sustainable cultivation practices.Keywords: Weed detection, Onion crop, Convolutional Neural Networks (CNN), Precision agriculture, Transfer learning, Image processing.