Toxic Plant Classification Using Convolutional Neural Networks

Toxic plants pose significant risks to both humans and animals, making their identification crucial for public health and environmental awareness.Many toxic species resemble non-toxic plants, leading to accidental ingestion or contact, which can result in severe health issues.Traditional identification methods rely on expert knowledge and manual classification, which can be time-consuming and prone to errors.This study presents an advanced toxic plant classification system using deep learning techniques, specifically leveraging the EfficientNet-B7 convolutional neural network (CNN) archi-tecture.The model is trained on a dataset consisting of 10,000 images, equally distributed between toxic and non-toxic plant species.Data augmentation techniques such as rotation, flip-ping, and contrast enhancement are applied to improve the model's generalization.EfficientNet-B7 is chosen for its superior balance of accuracy and computational efficiency, allowing for precise classification.The model is optimized using the AdamW optimizer with a learning rate scheduler, achieving a training accuracy of 99.48% and a validation accuracy of 97.89%.To enhance accessibility, the trained model is deployed using a user-friendly Streamlit-based web interface.Users can upload plant images for real-time classification, receiving instant feedback on the plant's toxicity status.The system also provides safety recommendations and treatment measures in case of exposure to toxic plants.The results demonstrate the effectiveness of the model, with minimal false positives and false negatives, making it a reliable tool for botanists, researchers, and outdoor enthusiasts.This research significantly contributes to plant identification and public safety by integrating cuttingedge deep learning methodologies with real-world applications.Future work in-cludes optimizing the model for mobile applications using Ten-sorFlow Lite, expanding the dataset for better generalization, and integrating additional plant species to enhance classification accuracy further.