基于机器视觉的黄瓜等级自动判别研究
Automatic Identification of Cucumber Level Based on Machine Vision
付彤 1雷金辉1
作者信息
- 1. 昆明理工大学信息工程与自动化学院，云南昆明650500
- 折叠
摘要
通过机器视觉研究黄瓜的形态特性，建立黄瓜量化分级评判标准，并总结出提取黄瓜特征值的理论算法。用Matalb软件实时采集读取CCD摄像机视频图像，再利用Matlab图像处理功能对采集到的图像进行图像滤波、图像分割处理，提取基于数学理论的表征黄瓜等级的特征值，建立人工神经网络的等级判别模型，训练神经网络，并进行泛化检测。结果表明，该等级判别模型的精度高达91．7％，具有一定的市场价值。
Abstract
Based on machine vision, morphological characteristics of cucumber were researched .Cucumber quantitative classification criteria were established.Theoretical algorithm of cucumber extraction eigenvalue was summarized.Matalb software was used to collect and read CCD camera video image.Image filtering and image segmentation of collected images was carried out by Matalb image processing function .Eigenvalue characterized the cucumber level was extracted based on mathematical theory .Level discriminant model of artificial neural network was estab-lished.The neural network training and testing were carried out.Results showed that the precision degree reached 91.7% and the model was of certain market value.
关键词
特征值/CCD摄像机/图像分割/人工神经网络Key words
Eigenvalues/CCD camera/Image segmentation/Artificial Neural Networks引用本文复制引用
出版年
2016
国家科技期刊平台
NSTL
维普
万方数据