基于人工神经网络与VPMCD的葡萄干等级检测方法研究

Research on raisin grade detection method based on artificial neural network and VPMCD

为了准确鉴别葡萄干等级,提出一种基于人工神经网络和VPMCD的葡萄干等级检测新方法。以新疆绿无核三个等级的葡萄干作为研究对象,提取颜色、形状的特征参数。采用BP神经网络算法,对比各特征组合对识别率的影响,确定了识别率较高的4个特征参数组合。最后应用VPMCD方法进行样本训练并进行葡萄干等级检测。将提出的方法与SVM、BP神经网络识别结果进行对比分析,结果表明,VPMCD算法识别率达到100%,分级效果明显,运算时间少,识别精度高,为农产品等级检测提供了一个新途径。

To precisely identify the raisin grades,a new raisin grade detection method based on artificial neural network and VPMCD is proposed. The Xinjiang Green seedless raisins of three grades are taken as the research object to extract the characteristic parameters of color and size. The BP neural network algorithm is used to compare the influence of each feature combination on identification rate. The four characteristic parameter combinations with high identification rate were determined. The VPMCD method is adopted to train the sample and detect the raisin grade. The identification result of the proposed method was compared with those of SVM method and BP neural network method. The comparison results show that the identification rate of VPMCD algorithm can reach up to 100%,and has superior classification effect,less operation time and high identification precision. It provides a new approach for grade detection of agricultural products.