International Research Journal of Education and Technology

Plant leaf diseases pose a major challenge in agriculture, reducing crop yield and quality. Early and accurate detection of these diseases is essential for implementing timely interventions, and this project focuses on developing an automated system for identifying and classifying plant leaf diseases using deep learning and computer vision techniques. The goal is to create an effective, cost-efficient solution that relies on high classification accuracy to distinguish various diseases from images of infected leaves. The system utilizes the VGG16 and LSTM models, both known for their strong performance in handling complex datasets. VGG16, a convolutional neural network, is well-suited for image classification due to its 16-layer architecture, capable of capturing intricate features from plant leaf images. Its deep layers are particularly effective at recognizing complex patterns and unique characteristics of specific plant leaf diseases. By leveraging VGG16 as a feature extractor, the system can learn to detect subtle variations and details in infected leaves, aiding in disease classification with high accuracy. LSTM, or Long Short-Term Memory, is a type of recurrent neural network designed to handle sequence-based data. If the dataset includes images showing disease progression stages over time, LSTM can be incorporated to analyze these sequences, making it possible to recognize patterns that develop across multiple stages of infection. This capability can provide insights into how the disease progresses, potentially allowing for a more comprehensive assessment of disease severity and helping users take preventive action before widespread damage occurs. In terms of functionality, the system communicates results via an Arduino cable and triggers a buzzer to alert users when a defective leaf is detected. This hardware integration provides real-time feedback, which is crucial for practical use in agricultural settings. By comparing the effectiveness of the VGG16 and LSTM approaches, the system can determine which model is best suited for plant leaf disease detection, taking into account both accuracy and computational efficiency.

Keywords: IoT, Mobile Safety, Ultrasonic Sensor, NodeMCU, Obstacle Detection, Vibration Motor, Real-Time Feedback, Accident Prevention, Wearable Devices, Mobile App Integration.