基于栈式稀疏自编码多特征融合的快速手势识别方法
|
摘要
针对复杂背景下手势分割提取效果不佳、图像识别率不高、识别困难等问题,研究多特征融合的快速手势识别方法.利用YCbCr颜色空间模型,构建肤色分布模型,从复杂背景中去除大部分非肤色的干扰,从而实现手势分割;接着采用5层栈式稀疏自编码网络框架,分别提取手势感兴趣区域(region of interest,ROI)的纹理图像、形状图像和显著视觉图像作为自编码网络输入,将提取到的不同类型的特征进行线性融合;最后使用基于径向基核函数(radial basis function,RBF)的支持向量机(support vector machine,SVM)分类器进行融合特征数据分类,从而实现不同类型的手势识别.实验结果表明,相比其他手势识别方法,本文方法识别率较高,提取特征更具有代表性,平均识别率可达95.05%.
In order to solve the problem of poor gesture segmentation extraction, low image recognition rate and difficult recognition, a multi-feature fusion method was studied for fast gesture recognition. Firstly, a skin color distribution model was established based on the YCbCr color space model, removing the most of non-skin color interference from the complex background, so as to realize the gesture segmentation. And then, taking the texture image of the gesture ROI(region of interest), the shape image and the significant visual image as self-coding network input, the different types of features were linearly merged according as 5-layer stacked sparse autoencoders network framework. Finally, a SVM(support vector machine) classifier based on RBF(radial basis function) kernel function was used to classify the characteristic data, so as to realize the gesture recognition for different types of gestures. Experimental results show that, compared with other gesture recognition methods, the recognition rate is higher and the extraction characteristics are more representative. The average recognition rate can reach 95.05%.
In order to solve the problem of poor gesture segmentation extraction, low image recognition rate and difficult recognition, a multi-feature fusion method was studied for fast gesture recognition. Firstly, a skin color distribution model was established based on the YCbCr color space model, removing the most of non-skin color interference from the complex background, so as to realize the gesture segmentation. And then, taking the texture image of the gesture ROI(region of interest), the shape image and the significant visual image as self-coding network input, the different types of features were linearly merged according as 5-layer stacked sparse autoencoders network framework. Finally, a SVM(support vector machine) classifier based on RBF(radial basis function) kernel function was used to classify the characteristic data, so as to realize the gesture recognition for different types of gestures. Experimental results show that, compared with other gesture recognition methods, the recognition rate is higher and the extraction characteristics are more representative. The average recognition rate can reach 95.05%.
引文
[1] Jevtic A,Doisy G,Parmet Y,et al.Comparison of interaction modalities for mobile indoor robot guidance:direct physical interaction,person following,and pointing control[J].IEEE Transactions on Human-Machine Systems,2015,45(6):653-663.
[2] Cheng H,Yang L,Liu Z C.Survey on 3D hand gesture recognition[J].IEEE Transactions on Circuits and Systems for Video Technology,2016,26(9):1659-1673.
[3] Kawulok M,Nalepa J,Kawulok J.Skin detection and segmentation in color images[M]//Advances in Low-Level Color Image Processing.Netherlands:Springer,2014:329-366.
[4] Weng C,Li Y,Zhang M,et al.Robust hand posture recognition integrating multi-cue hand tracking[C]//Proceedings of International Conference on E-Learning and Games.Changchun,China:IEEE,2010:497-508.
[5] 林海波,王圣彬,张毅.基于改进深度信息的手势分割与定位[J].计算机应用,2017,37(1):251-254.Lin Haibo,Wang Shengbin,Zhang Yi.Gesture segmentation and positioning based on improved depth information[J].Journal of Computer Applications,2017,37(1):251-254.(in Chinese)
[6] Gamal M,Abdul K,Sallam E.Hand gesture recognition using fourier descriptors[C]//Proceedings of International Conference on Computer Engineering and Systems.Cairo,Egypt:IEEE,2013:274- 279.
[7] Phung S L,Bouzerdoum A,Chai D.A novel skin color model in YCbCr color space and its application to human face detection[C]//Proceedings of International Conference on Image Processing.[S.l.]:IEEE,2002:289 -292.
[8] Itti L,Koch C,Niebur E.A model of saliency-based visual attention for rapid scene analysis[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1998,20(11):1254-1259.
[9] Sch?lkopf B,Platt J,Hofmann T.Graph-based visual saliency[C]//Proceedings of Neural Information Processing Systems.[S.l.]:IEEE,2006:545-552.
[10] Goodfellow I J,Courville A C,Bengio Y,et al.Joint training deep Boltzmann machines for classification[J].arXiv:Machine Learning,2013,25(3):327-335.
[11] Tang C,Ou Y,Jiang G,et al.Hand tracking and pose recognition via depth and color information[C]//Proceedings of IEEE International Conference on Robotics and Biomimetics.[S.l.]:IEEE,2012:1104-1109.
[12] Li Y.Hand gesture recognition using Kinect[C]//Proceedings of IEEE International Conference on Software Engineering and Service Science.[S.l.]:IEEE,2012:196-199.
[13] 赵爱芳,裴东,王全州,等.复杂环境中多信息融合的手势识别[J].计算机工程与应用,2014,50(5):180-184.Zhao Aifang,Pei Dong,Wang Quanzhou,et al.Gesture recognition fused with multi-information in complex environment[J].Computer Engineering and Applications,2014,50(5):180-184.(in Chinese)
[2] Cheng H,Yang L,Liu Z C.Survey on 3D hand gesture recognition[J].IEEE Transactions on Circuits and Systems for Video Technology,2016,26(9):1659-1673.
[3] Kawulok M,Nalepa J,Kawulok J.Skin detection and segmentation in color images[M]//Advances in Low-Level Color Image Processing.Netherlands:Springer,2014:329-366.
[4] Weng C,Li Y,Zhang M,et al.Robust hand posture recognition integrating multi-cue hand tracking[C]//Proceedings of International Conference on E-Learning and Games.Changchun,China:IEEE,2010:497-508.
[5] 林海波,王圣彬,张毅.基于改进深度信息的手势分割与定位[J].计算机应用,2017,37(1):251-254.Lin Haibo,Wang Shengbin,Zhang Yi.Gesture segmentation and positioning based on improved depth information[J].Journal of Computer Applications,2017,37(1):251-254.(in Chinese)
[6] Gamal M,Abdul K,Sallam E.Hand gesture recognition using fourier descriptors[C]//Proceedings of International Conference on Computer Engineering and Systems.Cairo,Egypt:IEEE,2013:274- 279.
[7] Phung S L,Bouzerdoum A,Chai D.A novel skin color model in YCbCr color space and its application to human face detection[C]//Proceedings of International Conference on Image Processing.[S.l.]:IEEE,2002:289 -292.
[8] Itti L,Koch C,Niebur E.A model of saliency-based visual attention for rapid scene analysis[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1998,20(11):1254-1259.
[9] Sch?lkopf B,Platt J,Hofmann T.Graph-based visual saliency[C]//Proceedings of Neural Information Processing Systems.[S.l.]:IEEE,2006:545-552.
[10] Goodfellow I J,Courville A C,Bengio Y,et al.Joint training deep Boltzmann machines for classification[J].arXiv:Machine Learning,2013,25(3):327-335.
[11] Tang C,Ou Y,Jiang G,et al.Hand tracking and pose recognition via depth and color information[C]//Proceedings of IEEE International Conference on Robotics and Biomimetics.[S.l.]:IEEE,2012:1104-1109.
[12] Li Y.Hand gesture recognition using Kinect[C]//Proceedings of IEEE International Conference on Software Engineering and Service Science.[S.l.]:IEEE,2012:196-199.
[13] 赵爱芳,裴东,王全州,等.复杂环境中多信息融合的手势识别[J].计算机工程与应用,2014,50(5):180-184.Zhao Aifang,Pei Dong,Wang Quanzhou,et al.Gesture recognition fused with multi-information in complex environment[J].Computer Engineering and Applications,2014,50(5):180-184.(in Chinese)