当前位置: 首页 > 文章 > 基于C-SVM的碎米检测算法研究 湖北农业科学 2016,55 (20) 5368-5371
Position: Home > Articles > The Algorithm Research of Broken Rice Detection Based on C-SVM Hubei Agricultural Sciences 2016,55 (20) 5368-5371

基于C-SVM的碎米检测算法研究

作  者:
梁诗华;林毅鑫;何金成
单  位:
福建农林大学机电工程学院
关键词:
碎米;特征提取;Orange Canvas;C-SVM
摘  要:
提出了一种基于支持向量机(C-SVM)区分整精米和碎米的方法,首先对大米图像进行阈值分割、平滑处理等图像预处理,并根据大米的粒形特点,提取米粒的面积、周长等6个形态特征,利用Orange Canvas数据挖掘软件对C-SVM中核函数参数进行预判,最终选择线性核函数的C-SVM作为分类器进行分类。对8组大米样本图像进行碎米测试,可达到较好的分类效果。试验结果表明,线性核函数的支持向量机对精整米与碎米识别分类准确率为95.6%。
译  名:
The Algorithm Research of Broken Rice Detection Based on C-SVM
作  者:
LIANG Shi-hua;LIN Yi-xin;HE Jin-cheng;College of Mechanical and Electrical Engineering , Fujian Agriculture and Forestry University;
关键词:
broken rice;;feature extraction;;Orange Canvas;;C-SVM
摘  要:
A method based on support vector machine(C-SVM) to distinguishbetween head rice and broken rice was proposed. Firstly,it did the image threshold segmentation,then proceeded the smooth processing. And according to the characteristics of rice grain shape,extracted six morphological characteristics such as area,perimeter and so on. Then using Orange Canvas data mining software to predict kernel function parameter. Finally,linear kernel function was selected as classifier.Eight groups of broken rice samples had been tested,which achieved a preferable classification result. The test results showed that linear of SVM can identify head rice and broken rice with classification accuracy at 95.6%.

相似文章

计量
文章访问数: 18
HTML全文浏览量: 0
PDF下载量: 0

所属期刊

推荐期刊