Research on Quantification Method of Maize Leaf Phenotype Parameters Based on Machine Vision

Abstract

This paper aims to obtain a model (method) for monitering the growth of corn automatically and accurately. The latest computer vision technology is applied to calculate the method of extracting the surface characteristic parameters. Binocular vision technology is used to calculate the 3D point cloud of maize image. Meanwhile, the 3D point noise reduction established by bilateral filtering is adopted to establish the 3D reconstruction model of maize. In the 3D model, the height and width of maize are calculated proportionally. The error rate of the proposed method is 0.52%. This work provides a reference for the growth monitoring and virtual growth of corn.