摘要
针对小麦品种分类的问题,通过对小麦籽粒投影面积、籽粒周长、籽粒紧度、籽粒长度、籽粒宽、不对称系数、籽粒槽长度等7个参数进行分析,建立数学模型对测试样本进行准确分类,分析数据设计小麦籽粒识别算法,识别三种不同小麦品种的籽粒.实施步骤为,首先构建小麦品种分类的数学模型,接着使用层次分析法构建了小麦品种的评价指标,使用混淆矩阵完成评价,最后通过卡玛、蔷薇、加拿大人三种小麦品种分类的实例分析,提出的算法能够优化数学模型,提高小麦品种的识别率.
Abstract
The paper establishes a mathematical model to accurately classify the test samples through parameters such as grain projection area,grain perimeter,grain compactness,grain length,grain width,asymmetry coefficient,grain trough length,and de-signs a mathematical algorithm for wheat grain recognition by analyzing the data to identify the grains of three different wheat varieties.First,the mathematical model for wheat variety classification is constructed,Then,the evaluation index of wheat varieties is construc-ted by using AHP,and the confusion matrix is used to evaluate successfully.Finally,the mathematical model of wheat varieties is op-timized by practical analysis.Through the analysis of the classification of three wheat varieties,Kama,Rosa and Canadians,after ex-cluding parameters such as perimeter and grain projection area,the model can be optimized to the greatest extent and the recognition rate of wheat varieties can be improved.
维普
万方数据