An Intelligent IoT-Based Framework for Plant Disease Detection and Crop Advisory Using CNN and ESP32

Abstract

The increasing demand for sustainable agricultural production has created a need for intelligent systems that continuously monitor crop-growing conditions and support timely agricultural decision-making. Conventional farming practices largely depend on manual observation for monitoring environmental conditions and identifying plant diseases, which can be time-consuming and require expert knowledge. This paper presents an intelligent IoT-based plant disease monitoring and crop advisory system integrating an ESP32 microcontroller, environmental sensors, a Convolutional Neural Network (CNN), and a Flask-based web application. The system continuously monitors temperature, humidity, soil moisture, soil pH, water level, light intensity, and air quality. Sensor readings collected through the ESP32 are processed, stored, and displayed on a real-time web dashboard. Users can also upload plant leaf images, which are preprocessed and analyzed using a trained CNN model to identify disease categories based on visual characteristics such as colour, texture, and disease patterns. The system provides disease information and preventive recommendations. Furthermore, the crop advisory module compares real-time sensor values with predefined crop requirements and generates recommendations for irrigation, soil moisture, pH management, and crop maintenance. The integrated framework reduces dependence on manual inspection and supports data-driven agricultural decisions, providing a practical, scalable, and cost-effective approach to smart agriculture

The increasing demand for sustainable agricultural production has created a need for intelligent systems that continuously monitor crop-growing conditions and support timely agricultural decision-making. Conventional farming practices largely depend on manual observation for monitoring environmental conditions and identifying plant diseases, which can be time-consuming and require expert knowledge. This paper presents an intelligent IoT-based plant disease monitoring and crop advisory system integrating an ESP32 microcontroller, environmental sensors, a Convolutional Neural Network (CNN), and a Flask-based web application. The system continuously monitors temperature, humidity, soil moisture, soil pH, water level, light intensity, and air quality. Sensor readings collected through the ESP32 are processed, stored, and displayed on a real-time web dashboard. Users can also upload plant leaf images, which are preprocessed and analyzed using a trained CNN model to identify disease categories based on visual characteristics such as colour, texture, and disease patterns. The system provides disease information and preventive recommendations. Furthermore, the crop advisory module compares real-time sensor values with predefined crop requirements and generates recommendations for irrigation, soil moisture, pH management, and crop maintenance. The integrated framework reduces dependence on manual inspection and supports data-driven agricultural decisions, providing a practical, scalable, and cost-effective approach to smart agriculture