机器视觉在花椒品种鉴别中的初步研究

发布时间：2018-06-20 03:05

本文选题：机器视觉 + 颜色特征　；参考：《传感技术学报》2016年01期

【摘要】：花椒是一种重要的烹饪调料和中药配料。本文将机器视觉技术引入到花椒品种的快速鉴别中,通过机器视觉硬件装置获取6类花椒样品图像共90幅,其中60幅作为训练样本,30幅作为测试样本。对所有样本分别提取了颜色和纹理的共10个特征参数,利用概率神经网络对特征数据进行鉴别,正确识别率为93.33%。本文研究的基于机器视觉的花椒品种鉴别方法可以快速准确地提取花椒样品的特征数据,为批量分选花椒奠定了技术基础。
[Abstract]:Zanthoxylum bungeanum is an important cooking seasoning and traditional Chinese medicine ingredients. In this paper, the machine vision technology is introduced into the rapid identification of pepper varieties. A total of 90 images of 6 kinds of pepper samples are obtained by using the machine vision hardware device, of which 60 samples are used as training samples and 30 samples are used as test samples. A total of 10 feature parameters of color and texture were extracted from all samples, and the probability neural network was used to identify the feature data. The correct recognition rate was 93.33. The method based on machine vision can quickly and accurately extract the characteristic data of Chinese prickly ash samples and lay a technical foundation for batch sorting of Chinese prickly ash.
【作者单位】：河南农业大学理学院;
【基金】：河南省科技攻关计划项目(142102110054) 河南省教育厅科学技术研究重点项目(12A510014)
【分类号】：S573.9;TP391.41

【相似文献】