Estimation of pomegranate shape and size by machine vision

Abstract

A simplified machine vision system for estimating shape and size of pomegranates was developed. The system allows the estimation ofvolume, surface area, and weight ofthe fruit using predic-tion equations developed from the relationship between projected area and shape and size. This proce-dure resulted in excellent values of the coefficient of determination and can be used to estimate volume, surface area, and weight of Banati and Manfaluti pomegranates. The analysis of variance indicated that there was no significant difference in shape between Banati and Manfaluti cultivars at the 95% confidence level. A power law equation was obtained to describe the relationship between volume, surface area, and weight and the projected areas. The average measurement error of prediction was in the 1.5. 0.7 and 2.0% - range for volume, surface area, and weight of Banati pomegranate, respectively:'The Maiif'aluti pomegra-nate had average measurement errors in the range of 1.1, 1.4 and 3.2% for volume, surface area, and weight, respectively.