Assessment of ‘hen and chicken’disorder for marketable yield estimates of table grape using the ‘Berry Analysis Tool’

Abstract

Bunch and berry size are important quality criteria for table grapes, and bunch weight is used in estimation of vine yield. The Berry Analysis Tool (BAT), a machine vision technique, was trialled for use in estimation of berry count, diameter and weight, to support estimates of ‘hen and chicken’ disorder and vineyard yield. Berries were spread on a plate for imaging. BAT estimates achieved a r2 = 0.99 on berry number per bunch, r2 = 0.98 on average berry minor axis and r2 = 0.99 on bunch weight. Based on an allometric relationship between lineal dimensions and weight, these attributes were used to estimate the proportion of under-size (‘chicken’) berries per bunch, and bunch weight. The estimated bunch weight multiplied by a number bunches per vine provides an estimation of vine yield. Use of the BAT as a decision support tool in table grape farm management is described.