Apple yield mapping using hyperspectral machine vision

For orchard growers, it is important to estimate the quantity of fruit on the trees at different stages of their growth. This study proposes a method of automatically detecting apples in digital images that can be used for automating the yield estimation of apples on trees at different stages of their growth by means of machine vision. This investigation concentrates on estimating yield of green varieties of apples. To achieve this goal, hyperspectral imaging was applied. A multistage algorithm was developed which utilizes PCA and ECHO as well as machine vision techniques. The overall correct detection rate was 87.0% with an overall error rate of 14.9%.