Prediction of Sugar and Carbohydrate in Sapota using Hyperspectral Image Segmentation

India offers a diverse range of fruits to its consumers. Each fruit has distinct characteristics, physical and chemical, defining their ripening and consumption stages. Fruits that lack strong physical attribute changes must be chemically analyzed for their ripeness and consumption. One such fruit is sapota, a tropical evergreen fruit-bearing tree. Sapota, a delicious fruit, possesses numerous medicinal and nutritive properties. It has a shorter shelf life, hence, is prone to major post-harvest losses. To overcome this difficulty, we propose a model where we use hyperspectral image segmentation to predict the sugar and carbohydrate values of the fruit. This model determines the sugar and carbohydrate values and in turn, predicts the ripeness of the fruit as there is an increase in the sugar and carbohydrate contents in the fruit with its ripening. The accurate values from the laboratory tests are used to test our model. A near-infrared Hyperspectral Imaging (HSI) system scanned samples with a wavelength range of 874-1734 nm. This model comes as an asset at the industry level and helps determine the ripeness of a large batch of sapota for its timely storage and relegation to retailers and consumers.