Optimizing Plant Health: A Deep Learning Ensemble Approach to Nutrient Deficiency Detection

In this paper, a complete multimodal architecture to predict various plant nutrient deficiencies, based on deep learning techniques, is reported. Integrated diverse data from image-based models (LeNet, MobileNetV1, AlexNet, ResNet, GoogleNet) symptoms-based text and audio to provide precise and comprehensive deficiency detection Using visual conditions from plant photos with text or audio descriptions on what the plant is experiencing, the system guarantees accurate predictions on different interpretations. National Center for Agricultural Research and Extension, the system provides personalized treatment suggestions, allowing farmers to make timely and effective decisions about how to manage nutrients. The experimental results show that the framework obtains high accuracy and flexibility in real life agriculture environments, which is promising for crop health monitoring and sustainable agriculture.