Apple Fruit Disease Detection for Hydroponic Plants Using Machine Learning and Image Processing

Abstract

Agriculture played a major role in the economy of many developing countries. Apple falls under the commercial crops as a fruit that will be grown worldwide. In the whole growth cycle of apples, there are many types of apple diseases and pests, therefore, recognition and determination of these illnesses are essential. K-means Cluster segmentation is used to detect the diseases in the apple fruit. In the feature extraction method, the Threshold segmentation process is applied and extracts the pixels from the apple and predicts the disease present in apple fruit. The machine learning algorithm like, Support Vector Machine and Deep learning algorithm like Convolutional Neural Network algorithms are used to test & train the data. The tool used for implementation is Anaconda Spyder 3.7 and Python Language. It is free of cost and the most popular non-proprietary dispensation software of Python programming. The dataset contains both healthy and unhealthy apple fruit images and a confusion matrix for validating the proposed system will change depending on the dataset. The outcomes visualize the higher arrangement exactness of the proposed SVM and CNN model. At last, the outcome will be generated in the form of loss, accuracy, precision, recall, sensitivity, and specificity. Keywords— : K-means Clustering, Support Vector Machine, Classification, Classification Neural Network