Diagnosing Nutrient Deficiency Symptoms in Citrus with Artificial Neural Networks.

Visual identification of nutrient deficiencies in foliage is an important diagnostic tool for fine-tuning nutrient management of citrus. Optimal nutrition is critical to to maximize growth, fruit yields and quality. In this study, a deep learning convolutional neural network model was trained to recognize deficiency symptoms of iron, zinc, magnesium, manganese, and potassium on citrus leaves. The performance of the training, internal and external validation phases of the model will be discussed and compared with the performance of human diagnostics.