Protecting the environment from pollution through early detection of infections on crops using the deep belief network in paddy

Paddy is the staple food for more than 50% of 138 billion Indian population. Inorder to meet with the growing demand, farmers often resort to application of synthetic fertilizers and plant protection chemicals indiscriminately. Rice is susceptible to diseases, pests, and nutrient deficiencies likewise other crops. Ignorant about the reasons for damage farmers apply synthetic chemicals that too in exorbitant rates. Excessive use of these chemical molecules alters the soil characteristics and causes environmental pollution as well. As a result, entire eco system gets affected. To overcome this, it is necessary to identify the reason for damage early and necessary treatments should be done in the beginning stages itself. Early detection can be done by assessing the leaves and culm of paddy. Assessment by naked eye may misinterpret symptoms and if artificial intelligence is used such misinterpretations can be minimised. This study proposes an automatic classification system using artificial intelligence and image processing for identification of diseased, pest infested and nutrient deficient crop using symptoms exhibited in the leaves and culm of paddy. Kaggle data set was being used to test the performance of the proposed classification system for metrics specificity, precision, sensitivity, F1-score and accuracy. The proposed work provides a specificity, precision, sensitivity, F1-score and accuracy of 97.1%, 97.6%, 96.2%, 96.8%, and 98.1% respectively. The evaluation results indicate that the proposed algorithm outperforms other recent rice leaf disease, pest and nutrient deficiency classification algorithms. Thus, precise identification of reasons for infection allows farmers to use specific control methods with less toxic chemicals or through eco-friendly methods. Thus, environmental pollution and soil characteristics can be saved and in turn can save the environment and its creatures.