Volume estimation of strawberries, mushrooms, and tomatoes with a machine vision system

Abstract

A computer vision technique was used to determine the volume of raw agricultural products with an irregular shape. Thirty images of each rotated raw roma tomato, salad tomato, white button mushroom, and strawberry were collected. Volume determinations of these products using the optical imaging system were compared to volume measurements collected using a water displacement/buoyant force method. A high correlation was found between measurements from both methods. Furthermore, regression analyses of weight and volume measurements of each product set were used to develop equations to predict object volume from object weight. Weight measurement of raw produce can be used as a non-destructive method to estimate unit volume for sorting and grading purposes.