Modeling of Kiwifruit Mass Based on Dimensions and Projected Areas
Fruits are often graded on the basis of size, but it may be more suitable and economical to develop a system which grades by mass. Therefore, a relationship between mass and size of fruits is needed. In this study, eighth models for predicting mass of kiwifruit (Actinidia deliciosa) from its dimensions and projected areas were identified. Models were divided into two classifications: 1- Single and multiple variable regressions of kiwifruit dimensions (1 classification models). 2- Single and multiple variable regressions of kiwifruit st projected areas (2 classification models). The most common commercial variety of kiwifruit cv. Hayward was nd selected for this study. About 125 samples of kiwifruit were picked up at random and physical attributes of kiwifruit such as mass, dimensions and projected areas were measured. The results of the study indicated that second classification models had higher performance. However, second classification models need electronic systems with cameras, whereas 1 classification models, except multiple variable one, are used in the simple st mechanical systems. The results of the study also indicated that in order to predict mass of kiwifruit based on dimensions, the mass model based on the major diameter as M = 2.19 a – 59.36 with R = 0.74 is recommended. 2 Also, to predict mass of kiwifruit based on projected areas, the mass model based on the first projected area as M = 4.06 PA – 19.35 with R = 0.94 can be suggested.