Abstract
In this study we developed a technique to estimate acorn yield using airborne Multi-spectral imaging. All models were experimentally validated by a leave-one-out cross validation technique. On the airborne Multi-spectral image, the canopies of 22 individual acorn trees were identified. The mean values of the reflection radiance, tree size, and canopy area were extracted from the images as independent variables. Cup-shaped cupules were collected using seedtraps, and the number of cupules, after applying logarithmic transformation was considered as the dependent variable. Results indicated that the images had significant relationships (2003: R2=0.49, 2004: R2=0.56), and these relationships of stability were confirmed in the cross validation.
