A deep learning approach for lantana camara weed detection and localization in the natural environment

This paper presents the development of a deep learning approach for detecting and localizing lantana weed in the natural environment. Based on the YOLO framework, object detection in images is accomplished through convolutional neural networks. By using the DeepWeed dataset which was collected from eight locations across northern Australia, we have trained several neural networks to detect the presence of lantana images. We found that the proposed approach outperformed the other algorithms in terms of accuracy in detecting the lantana weed, with an accuracy of 90.52%, a precision of 95.15%, and an F1-score of 91.69%.