Research Article

On-the-Go Grapevine Yield Estimation Using Image Analysis and Boolean Model

Table 2

Results obtained for the estimation of the berry number per cluster using the naïve estimator and the Boolean model in manually acquired cluster images of four different grapevine varieties.

Grapevine varietyManual countingNaïve estimatorBoolean model
Mean berry numberNumber of clustersMean berry numberRMSER2Mean berry numberRMSER2

Cabernet Sauvignon68.41526.748.80.3479.114.80.86
Grenache53.81218.237.50.7244.123.30.54
Syrah79.01132.350.70.7292.521.20.69
Tempranillo141.4743.9107.60.81136.622.00.66
Total85.74530.360.00.7188.120.10.80

Manual counting refers to the berry number obtained by manually destemming the cluster in the laboratory. The naïve and Boolean estimation was generated based on the analysis of the cluster images manually taken under field conditions. Asterisks represent statistical significance: ; ; .