基于机器视觉的脐橙可食率的研究

Study on Esculent Part of Navel Orange Based on Machine Vision

脐橙的可食率可通过脐橙的重量和体积估测得到,在此基础上提出了一种脐橙可食率的估测方法。首先通过机器视觉技术获取脐橙2个方向的截面积,再以这2个截面积和脐橙重量为输入变量的神经网络估计脐橙的可食率。试验结果表明,该方法在估计脐橙的可食率方面取得了较好的效果。

The percentage of the esculent part of naval orange could be estimated with weight and volume of navel orange,based on which a new estimation method is proposed.At first,the section areas from two directions of navel orange are obtained by machine vision.And then the percentage of the esculent part of naval orange is estimated with Neural Network whose input variables are two surface areas and weight of navel orange.Results show that the method achieves a good effect in estimating the edible rate of naval orange.