Non-destructive estimation of tomato fruit properties by interactive consecutive model series

Abstract

Fruit shape is an important quality parameter, and such variables as fruit diameter, height, weight, cross-sectional area and volume are components affecting this feature. In particular, these properties are the most important parameters in industrial applications for fruit grading, in determining the conditions of optimum packing, in providing the most suitable transportation facilities, and in optimizing crop production strategies. In this investigation, mathematical models were devised which enable estimation of the cross-sectional area, weight and volume of the fruit by a non-destructive method in the field before harvest. The modelling process was carried out by means of data analysis approaches and interactive consecutive calculation series for the Bandita F1 tomato cultivar. The correlation between the measured and estimated cross-sectional area, weight and volume of the fruit were 0.9672, 0.9809 and 0.9684, respectively. Apart from this, the accuracy rates of the models proposed for the estimation of the cross-sectional area, weight and volume are 97.12%, 95.40% and 95.37% respectively. In addition, the performance and validity of the models are in the “very good” category according to the all three analyses of NS, RSR and PBIAS. These results indicated that the models proposed gave high rates of accurate results