Hinting Pipeline and Multivariate Regression CNN for Maize Kernel Counting on the Ear
This work addresses the labor-intensive task of maize kernel counting for agricultural productivity assessment, offering an incremental improvement over existing methods.
The paper tackles the problem of manually counting maize kernels on ears by proposing a hinting preprocessing pipeline and a multivariate CNN regressor, achieving a MAE of 34.4 and R2 of 0.74, outperforming manual estimates with MAE 35.38 and R2 0.72.
Maize is a highly nutritional cereal widely used for human and animal consumption and also as raw material by the biofuels industries. This highlights the importance of precisely quantifying the corn grain productivity in season, helping the commercialization process, operationalization, and critical decision-making. Considering the manual labor cost of counting maize kernels, we propose in this work a novel preprocessing pipeline named hinting that guides the attention of the model to the center of the corn kernels and enables a deep learning model to deliver better performance, given a picture of one side of the corn ear. Also, we propose a multivariate CNN regressor that outperforms single regression results. Experiments indicated that the proposed approach excels the current manual estimates, obtaining MAE of 34.4 and R2 of 0.74 against 35.38 and 0.72 for the manual estimate, respectively.