Abstract

This work proposes a unique technique for the identification of micro and macro nutrient deficits in the leaves of the Okra (Abelmoschus esculentus L) plant. The approach utilises machine learning to accomplish this detection. The purpose is to create a system that is both efficient and accurate for identifying nutrient shortages, which is essential for maximising crop output and quality. The procedure entails the gathering of photographs of leaves, the extraction of features, and the categorization of the images via the use of machine learning algorithms. Method: The suggested method starts with the collection of high-resolution photographs of the leaves of the okra plant that are displaying indications of nutritional shortages. Techniques from the field of image processing are used in order to improve contrast, eliminate noise, and separate respective leaf sections. An operation known as feature extraction is carried out in order to get pertinent information about the morphology, colour, and texture of the leaf. These features are then used as input to machine learning techniques such as support vector machines (SVM), random forests, or convolutional neural networks (CNNs). Result: In the process of identifying micro and macro nutritional deficits in the leaves of the okra plant, the machine learning technique that was created exhibits promising positive outcomes. The identification of nutrient deficiency symptoms displays good levels of accuracy, sensitivity, and specificity, according to the results of an evaluation conducted on a heterogeneous dataset consisting of leaf pictures with ground truth labels. In terms of both accuracy and efficiency, the approach shows superior performance compared to the conventional visual inspection methods. In addition to this, the method demonstrates excellent resistance to changes in the lighting conditions and the orientation of the leaves. Conclusion: In conclusion, the machine learning technique that was provided offers a dependable and automated solution for the identification of micro and macro nutritional deficits in the leaves of the okra plant. Because of its precision and effectiveness, the technology is appropriate for incorporation into agricultural operations. There is a possibility that its applicability in real-world settings might be improved by further validation on bigger datasets and field tests, which would make sustainable agricultural production and food security easier to achieve.