Machine Learning Applications for Precise Nutrient Deficiency Detection in Paddy Farming Using K-Means Clustering and SVM

This research paper is focused on accurate detection of nutrient deficiencies in paddy crops through their leaves by application of machine learning techniques. Traditional methods for detecting nutrient deficiencies are considered to be visual observations which are labor-intensive with delayed precisions on the result. Thus, nutrient deficiencies in paddy crops are a challenging subject to consider for the research work with an advancement of machine learning and data-driven technologies to find an optimum solution to address the problem of the research work. The K-means clustering algorithm is used by the model for a sustainable development that includes various attributes and parameters for a better accuracy and precision. Few parameters can be described as feature selection, image processing, RGB to HSI conversion, and image segmentation along with performance analysis which uses F1 score, precision, and recall as indictors during evaluation. This model is validated using datasets from International Rice Research Institute (IRRI) and is able to demonstrate with an average accuracy of 85.3% to identify the nutrient deficiencies based on extracted features. This research will definitely contribute to agricultural field by offering an efficient and reliable system for the detection of nutrient deficiencies in paddy crops through their leaves.