Identification of pineapple maturity utilizing digital image using hybrid machine learning method Available to Purchase

Indonesia is the largest pineapple exporter in the world. Pineapple (Ananas comosus L. Merr) is a horticultural plant that is in demand by the public because it is rich in benefits at a low price. However, pineapple has characteristics that continue the function of metabolism. This metabolic process causes the pineapple fruit to ripen, age (senescence), and be damaged. Therefore, time is a very influential factor in the quality of pineapple. The accuracy and speed of sorting activities in postharvest are important things that must be considered. Good postharvest handling will reduce losses, both in quality and quantity. The longer the period between harvest and consumption, the greater the loss in quality. Identifying objects by relying on eye observation is still mostly done by farmers. Such a manual method may have shortcomings due to the limitations and weaknesses of the human sense of sight. In addition, the manual method takes a long time. So it is necessary to develop an intelligent system that can help speed up the sorting and grading process in postharvest activities. The identification and classification process uses digital images with color features (RGB and HSV), GLCM texture features, and shape features (area, major perimeter, and minor axis). While the machine learning methods used are SVM, KNN, and Neural Networks. The first thing to do is to collect the image of the pineapple, then preprocess it by cropping the unused part, then resizing and segmenting the image using the morphological method. After the 13-feature extraction process is complete, the next step is to identify the pineapple’s ripeness using the three methods. Based on the results of the study, it is known that the highest percentage of accuracy in the SVM method is 89%, the second neural network is 80%, and the last in the KNN method is 71%. By utilizing the image, it is expected to minimize the touch (friction) from the human hand, minimize the limitations (weaknesses) of the human sense of sight, and reduce the sorting time so that the quality of horticultural commodities will be maintained.