Abstract: Several studies have been made on identifying diseases in mulberry leaves, however, identifying nutrient deficiency in mulberry leaves has not been accomplished. The silkworms that feed on nutrient-deficient mulberry leaves produce low-quality silk. There is a great need for identifying nutrient-rich and healthy mulberry leaves for feeding the silkworm to get good quality silk yield. This paper is focused on segregating nutritious mulberry leaves for feeding the silkworms for cocoon formation. The process involves image acquisition, processing, segmentation, feature extraction, and classification. Auto-Encoder is used for feature extraction from mulberry leaves and for discrete them into nutritious and nutrient-deficient leaves. The real-valued feature vectors are passed to machine learning algorithms like the Naïve Bayes classifier algorithm, Support Vector Machine (SVM), and K-Nearest Neighbour (KNN) for classification. Among them KNN provides higher accuracy for segregating the leaves.
Keywords: Nutrient deficiency, Support Vector Machine (SVM), K-Nearest Neighbour (KNN), Naïve Bayes Classifier Algorithm, Auto-Encoder.
|
DOI:
10.17148/IJARCCE.2023.12132
