融合形状和颜色特征的苹果等级检测_李先锋.pdfVIP

202 2010 ，46 （35） Computer Engineering and Applications 计算机工程与应用 融合形状和颜色特征的苹果等级检测 1，2 1 2 2 李先锋 ，朱伟兴 ，花小朋 ，孔令东 1，2 1 2 2 LI Xian-feng ，ZHU Wei-xing ，HUA Xiao-peng ，KONG Ling-dong 1.江苏大学 电气信息工程学院，江苏 镇江 212013 2.盐城工学院 信息工程学院，江苏 盐城 224051 1.School of Electronic and Information Engineering ，Jiangsu University ，Zhenjiang ，Jiangsu 212013 ，China 2.School of Information Engineering ，Yancheng Institute of Technology ，Yancheng ，Jiangsu 224051 ，China E-mail ：vanguardlee@ LI Xian-feng ，ZHU Wei-xing ，HUA Xiao-peng ，et al.Apple grading detection based on fusion of shape and color features. Computer Engineering and Applications ，2010 ，46 （35 ）：202-204. Abstract ：In order to increase the accuracy and stability of apple grading ，shape and color features which can show the ap- ple ’s appearance quality are separately extracted by Fourier descriptor and HIS color model.Firstly ，the apples are graded re- spectively by neural network.Then ，the former grading results are used as evidences to achieve the decision fusion.Finally ，us- ing identification threshold to get the grades.The experimental results show that the grading accuracy reaches 93.75% ，the pro- posed method has good performance on accuracy and stability compared to the grading method based on single feature. Key words ：D-S evidence theory ；feature extraction ；Fourier descriptor ；HIS model ；decision fusion ；apple grading 摘 要：为了提高苹果分级的准确率和稳定性，在图像处理的基础上，基于Fourier 描述子和HIS 颜色模型分别提取了苹果的形状 和颜色两类主要外观特征，并分别用神经网络进行单特征初步分级，将其结果作为证据，通过D-S 证据理论进行决策级融合，根 据分类阈值得到最终分级结果。实验结果表明，该方法分级正确率达93.75% ，与单指标特征分级相比，识别率高，稳定性好。 关键词：D-S 证据理论；特征提取；傅里叶描述子；HIS 模型；决策级融合；苹果分级 DOI ：10.3778/j.issn. 1002-8331.2010.35.058 文章编号：1002-8331（2010 ）35-0202-03 文献标识码：Ａ 中图分类号：TP 18 1 引言 依据。假定基本概率分配（Basic Probab