AUTHOR=Liu Wentao , Wang Chenglin , Yan De , Chen Weilin , Luo Lufeng 

TITLE=Estimation of Characteristic Parameters of Grape Clusters Based on Point Cloud Data

JOURNAL=Frontiers in Plant Science

VOLUME=Volume 13 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2022.885167

DOI=10.3389/fpls.2022.885167

ISSN=1664-462X

ABSTRACT=The world's grape production is rich and manual picking is time-consuming. The development of intelligent grape picking has become possible. However, the individual differences of grapes pose a challenge to accurately measure their characteristic parameters. Therefore, this study explores a method for estimating the characteristic parameters of grapes based on the point cloud information obtained from the point cloud camera: Based on the segmentation of the grape point cloud by filtering and region growing algorithm, the 360°grape point cloud model is obtained by Principal Component Analysis (PCA) algorithm and iterative closest point (ICP) algorithm. After estimating model phenotypic size characteristics, five reconstruction methods were used to reconstruct grape models and the grape volumes under different reconstruction methods were calculated. Compare the volume measured by the drainage method with the volume estimated by the five algorithms: The grape characteristic parameters estimated by the combination of the point cloud processing algorithm and Poisson reconstruction algorithm are the closest to the measured grape characteristic parameters. The determination coefficient (R2) of the Poisson reconstruction algorithm is 0.9915, which is 0.2306 higher than that estimated by the alpha-shape algorithm (R2 = 0.7609). Therefore, the estimation method based on the point cloud proposed in this study can be applied to the estimation of characteristic parameters of grapes and provides a strong basis for grape quality assessment and intelligent picking strategies.