Detection of Plant Diseases in Hydroponics Farming Using Deep Learning Techniques

In a country like India, where most of the income depends on agriculture, it is challenging to rely only on soil-based agriculture in the future. These days, soil-based agriculture faces several difficulties, including the reckless use of pesticides and chemicals, decreasing the land's fertility, urbanization, natural disasters, and climate change. Numerous innovative agricultural techniques have emerged to overcome the drawbacks of conventional farming methods, among which hydroponic farming has appeared as a prominent and influential solution. Hydroponics farming is an innovative way to grow crops without using soil. Instead of planting seeds in dirt, hydroponics uses a nutrient-rich water solution to feed the plant's roots directly. In this work, images of diseased plant leaves caused by the deficiency of nitrogen, phosphorus, and potassium in the nutrient solution and healthy leaves were collected. After that, YOLOv8 active deep learning approach is proposed for detecting and classifying leaf disease caused by the deficiency of these nutrients.This proposed Yolov8 model achieves outstanding results in terms of evaluation metrics, namely accuracy, precision, F1-score, and recall, giving 99%, 99%, 98.7%, and 98.0%, respectively. These results represent the model's ability to classify and detect the infected plants accurately.