Predicting Health Status of Maize Crops by Integrating IoT Technology and Inception-v3 Convolutional Neural Network in Precision Agriculture

In Africa, particularly in Rwanda, maize serves as a significant income source. However, the growth of maize is often hampered by various diseases, resulting in decreased yield production. To mitigate these challenges and reduce human errors, farmers may take advantage of current emerging technologies such as artificial intelligence (AI) and Internet of things (IoT) for automating their farms by monitoring or regulating crucial resources like soil features, crops pests, and insects. The study specifically explores the utilization of an IoT-based system integrated with Inception-v3 of convolutional neural networks machine learning algorithm. The proposed solution incorporates a BME280 sensor to measure field pressure, humidity, and temperature in maize crops, along with an Inception-v3 convolutional neural network model to predict maize crop health status and provide real-time data insight to the farmers. The outcomes demonstrate exceptional performance, with an accuracy rate of 99.5%. Furthermore, the model has significantly reduced losses from 54 to 2%.