基于CLBP、改进KPCA和RF的牛肉大理石纹评级
|
摘要
为进一步提高牛肉大理石纹评级的正确率,提出了基于完整局部二值模式(Completed Local Binary Pattern,CLBP)、改进核主成分分析(Kernel Principal Component Analysis,KPCA)和随机森林(Random Forests,RF)的牛肉大理石纹评级方法。首先,利用CLBP提取牛肉大理石纹图像的纹理特征;其次,采用混沌蜂群算法对KPCA的核参数进行优化,使KPCA的降维效果和特征提取达到最优,获得表征牛肉大理石纹样本图像的特征向量;最后,使用随机森林完成牛肉大理石纹样本的分级识别,获得最终评级结果。大量实验结果表明,与基于分形维和图像特征的方法、基于灰度共生矩阵和BP(Back Propagation)神经网络法相比,本文方法所得识别率最高。
In order to further improve the correct rate of beef marbling grading, a beef marbling grading method based on completed local binary pattern(CLBP), kernel principal component analysis and random forests is proposed. Firstly, CLBP is used to extract texture features of beef marbling image. Then the kernel parameters of kernel principal component analysis is optimized by chaos artificial bee colony, which makes dimensionality reduction and feature extraction by KPCA to achieve the optimal effect. Thus feature vectors are obtained to characterize the beef marbling images. Finally, random forests is applied to complete classification recognition of beef marble samples and get the final ratings result. A large number of experimental results show that, compared with method based on fractal dimension and image features, method based on gray level co-occurrence matrix, and the method based on Back Propagation neural network, the proposed method attains the highest recognition rate.
In order to further improve the correct rate of beef marbling grading, a beef marbling grading method based on completed local binary pattern(CLBP), kernel principal component analysis and random forests is proposed. Firstly, CLBP is used to extract texture features of beef marbling image. Then the kernel parameters of kernel principal component analysis is optimized by chaos artificial bee colony, which makes dimensionality reduction and feature extraction by KPCA to achieve the optimal effect. Thus feature vectors are obtained to characterize the beef marbling images. Finally, random forests is applied to complete classification recognition of beef marble samples and get the final ratings result. A large number of experimental results show that, compared with method based on fractal dimension and image features, method based on gray level co-occurrence matrix, and the method based on Back Propagation neural network, the proposed method attains the highest recognition rate.
引文
[1]汤晓艳,王敏,钱永忠,等.牛肉分级标准及分级技术发展概况综述[J].食品科学,2011,32(19):288-293.
[2]周彤,彭彦昆.牛肉大理石花纹图像特征信息提取及自动分级方法[J].农业工程学报,2013,29(15):286-293.
[3]陈坤杰,姬长英.牛肉自动分级技术研究进展分析[J].农业机械学报,2006,37(3):153-156.
[4]TAN J.Meat quality evaluation by computer vision[J].Journal of Food Engineering,2004,61(1):27-35.
[5]FUKUDA O,NABEOKA N,MIYAJIMA T.Estimation of marbling score in live cattle based on ICA and a neural network[C].IEEE 2013 International Conference on Systems Man and Cybernetics,Manchester,2013:1622-1627.
[6]陈坤杰,姬长英.基于图像运算的牛肉大理石花纹分割方法[J].农业机械学报,2007,38(5):195-196.
[7]陈坤杰,吴贵茹,於海明,等.基于分形维和图像特征的牛肉大理石花纹等级判定模型[J].农业机械学报,2012,43(5):147-151.
[8]谢元澄,徐焕良,谢庄.基于牛肉大理石花纹标准(BMS)图像的纹理特征分析[J].中国农业科学,2010,43(24):5121-5128.
[9]张建勋,李涛,孙权,等.猪眼肌B超图像纹理特征提取与分类[J].重庆理工大学学报:自然科学版,2013(2):74-78.
[10]Guo Zhenhua,Zhang Lei,ZHANG D.A completed modeling of local binary pattern operator for texture classification[J].IEEE Transactions on Image Processing,2010,19(6):1657-1663.
[11]Zhang Yi,Han Jing,Yue Jiang,et al.Weighted KPCA degree of homogeneity amended non-classical receptive field inhibition model for salient contour extraction in low-light-level image[J].IEEE Transactions on Image Processing,2014,23(6):2732-2743.
[12]CHOI J H,SONG G Y,LEE J W.Road identification in monocular color images using random forest and color correlogram[J].International Journal of Automotive Technology,2012,13(6):941-948.
[13]杨帆,林琛,周绮凤,等.基于随机森林的潜在k近邻算法及其在基因表达数据分类中的应用[J].系统工程理论与实践,2012,32(4):815-825.
[14]陈超,李文藻.一种基于随机森林与颜色特征的岩屑识别算法[J].四川大学学报(自然科学版),2012,49(3):587-592.
[15]曹正凤.随机森林算法优化研究[D].北京:首都经济贸易大学,2014.
[2]周彤,彭彦昆.牛肉大理石花纹图像特征信息提取及自动分级方法[J].农业工程学报,2013,29(15):286-293.
[3]陈坤杰,姬长英.牛肉自动分级技术研究进展分析[J].农业机械学报,2006,37(3):153-156.
[4]TAN J.Meat quality evaluation by computer vision[J].Journal of Food Engineering,2004,61(1):27-35.
[5]FUKUDA O,NABEOKA N,MIYAJIMA T.Estimation of marbling score in live cattle based on ICA and a neural network[C].IEEE 2013 International Conference on Systems Man and Cybernetics,Manchester,2013:1622-1627.
[6]陈坤杰,姬长英.基于图像运算的牛肉大理石花纹分割方法[J].农业机械学报,2007,38(5):195-196.
[7]陈坤杰,吴贵茹,於海明,等.基于分形维和图像特征的牛肉大理石花纹等级判定模型[J].农业机械学报,2012,43(5):147-151.
[8]谢元澄,徐焕良,谢庄.基于牛肉大理石花纹标准(BMS)图像的纹理特征分析[J].中国农业科学,2010,43(24):5121-5128.
[9]张建勋,李涛,孙权,等.猪眼肌B超图像纹理特征提取与分类[J].重庆理工大学学报:自然科学版,2013(2):74-78.
[10]Guo Zhenhua,Zhang Lei,ZHANG D.A completed modeling of local binary pattern operator for texture classification[J].IEEE Transactions on Image Processing,2010,19(6):1657-1663.
[11]Zhang Yi,Han Jing,Yue Jiang,et al.Weighted KPCA degree of homogeneity amended non-classical receptive field inhibition model for salient contour extraction in low-light-level image[J].IEEE Transactions on Image Processing,2014,23(6):2732-2743.
[12]CHOI J H,SONG G Y,LEE J W.Road identification in monocular color images using random forest and color correlogram[J].International Journal of Automotive Technology,2012,13(6):941-948.
[13]杨帆,林琛,周绮凤,等.基于随机森林的潜在k近邻算法及其在基因表达数据分类中的应用[J].系统工程理论与实践,2012,32(4):815-825.
[14]陈超,李文藻.一种基于随机森林与颜色特征的岩屑识别算法[J].四川大学学报(自然科学版),2012,49(3):587-592.
[15]曹正凤.随机森林算法优化研究[D].北京:首都经济贸易大学,2014.