Prediction of volume, weight and surface area of banana (Musa acuminata) using picture image analysis

The design of processing equipments, materials and processes for the handling of crops require data of the physical properties of the crops such as weight, volume and surface area. Obtaining these data can be very tedious and destructive to the samples being used. This research aims at a non- destructive, fast, but accurate and easy way of finding volume, weight, and surface area of a crop from its picture image. Banana (Musa acuminata) was used for the study and the mean weight, volume, surface area, plan area, and plan perimeter were 142.07±23.48 g, 158.90±33.14 cm³, 171.25±22.32 cm², 79.34±10.45 cm² and 47.41±3.72 cm, respectively. The values obtained were correlated, using plan area and perimeter as the independent variables and weight, volume and surface area as the dependent variables. The prediction equations obtained using plan area measurements are Weight (g) = 1.866×plan area (cm²) - 5.9649; Volume (cm³) = 1.8518×plan area (cm²) - 11.978; Surface area (cm²) = 1.5568×plan area (cm²) - 47.737. These prediction equations offer better prediction values than perimeter, but of these equations only the prediction equation for weight can be used with an appreciable r-squared value of 68.89%. Therefore, this equation can be used by designers after obtaining a picture image of the banana.