基于多特征融合的指挥手势识别方法研究
|
摘要
针对飞行甲板指挥手势识别特定应用需求,提出一种基于多特征融合的手势识别方法。利用深度摄像头采集到的视觉信息,从轨迹和手形两方面特征入手,建立了三维轨迹特征向量和手形稀疏表示。一方面基于轨迹特征通过轨迹归一化、重采样、对齐等处理进行识别,另一方面基于HOG(Histogramof Oriented Gradients)特征通过稀疏观察对齐进行手形识别,将识别结果进行有效融合。实验结果表明,提出的基于多特征融合的指挥手势识别方法在准确率上有较大提升,同时具有较好的鲁棒性。
Aiming at the specific application requirements of command gesture for flight deck, a gesture recognition method based on multi-feature fusion is proposed. The 3 D trajectory feature vector and hand sparse representation are established from two aspects of the trajectory and posture based on the visual information collected by depth camera. On the one hand, the gesture is recognized through normalization resampling and alignment based on the trajectory feature. On the other hand, the gesture is recognized through sparse representation alignment based on the HOG feature. The recognition results are fused effectively. The experimental results indicate that our methods greatly enhance accuracy, and have better robustness.
Aiming at the specific application requirements of command gesture for flight deck, a gesture recognition method based on multi-feature fusion is proposed. The 3 D trajectory feature vector and hand sparse representation are established from two aspects of the trajectory and posture based on the visual information collected by depth camera. On the one hand, the gesture is recognized through normalization resampling and alignment based on the trajectory feature. On the other hand, the gesture is recognized through sparse representation alignment based on the HOG feature. The recognition results are fused effectively. The experimental results indicate that our methods greatly enhance accuracy, and have better robustness.
引文
[1]胡容兵,谢君,傅冰.飞行甲板安全评价指标体系研究[J].海军工程大学学报,2013,25(6):93-98.Hu Rongbing,Xie Jun,Fu Bing.Analysis of Safety Evaluation Index System of Flight Deck[J].Journal of Naval University of Engineering,2013,25(6):93-98.
[2]Mohammadreza Asghari Oskoei,Huosheng Hu.Myoelectric Control Systems-A survey[J].Biomedical Signal Process&Control(S1004-1699),2007,2(4):275-294.
[3]胡琼,秦磊,黄庆明.基于视觉的人体动作识别综述[J].计算机学报,2013,36(12):2512-2524.Hu Qiong,Qin Lei,Huang Qingming.A Survey on Visual Human Action Recognition[J].Chinese Journal of Computers,2013,36(12):2512-2524.
[4]Carlsson S,Sullivan J.Action Recognition by Shape Matching to Key Frames[C]//European Conference on Computer Vision(ECCV).Copenhagen,Denmark:IEEE,2002:263-270.
[5]Wang Xiaogang,Tieu K,Grimson E.Learning Semantic Scene Models by Trajectory Analysis[C]//Proc of the 9th Europen Conference on Computer Vision.Berlin:Springer,2006:110-123.
[6]Shotton J,Fitzgibbon A,Sharp T,et al.Real-time Human Pose Recognition in Parts from A Single Depth Image[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.USA:Colorado Springs,2011:1297-1304.
[7]Jiang Zhuo-Lin,Lin Zhe,Davis L S.Recognizing human actions by learning and matching shape-motion prototype trees[J].IEEE Transactions on Pattern Analysis and Machine Intelligence(S1062-8828),2012,34(3):533-547.
[8]刘鑫辰,傅慧源,马华东.基于RGB-D摄像头的实时手指跟踪与手势识别[J].计算机学报,2014,41(10):50-61.Liu Xinchen,Fu Huiyuan,Ma Huadong.Real-time Fingertip Tracking and Gesture Recognition Using RGB-DCamera[J].Computer Science,2014,41(10):50-61.
[9]Shi Qinfeng,Wang Li,Cheng Li,et al.Human action segmentation and recognition using discriminative semi-Markov Models[J].International Journal of Computer Vision(S0920-5691),2011,93(1):22-32.
[10]陶丽君,李翠华,张希婧,等.基于Kinect传感器深度信息的动态手势识别[J].厦门大学学报(自然科学版),2013,52(4):493-497.Tao Lijun,Li Cuihua,Zhang Xijing,et al.Dynamic Gesture Recognition Based on Kinect Depth Data[J].Journal of Xiamen University(Natural Science),2013,52(4):493-497.
[11]Datal N,Triggs B.Histograms of oriented grdients for human detection[C]//Proceedings of Conference on Computer Vision and Pattern Recognition,Montbonnot,France:IEEE,2005:886-893.
[12]张汉灵,李红英,周敏.融合多特征和压缩感知的手势识别[J].湖南大学学报(自然科学版),2013,40(3):87-92.Zhang Hanling,Li Hongying,Zhou Min.Hand Posture Recognition Based on Multi-feature and Compressive Sensing[J].Journal of Hunan University(Natural Sciences),2013,40(3):87-92.
[2]Mohammadreza Asghari Oskoei,Huosheng Hu.Myoelectric Control Systems-A survey[J].Biomedical Signal Process&Control(S1004-1699),2007,2(4):275-294.
[3]胡琼,秦磊,黄庆明.基于视觉的人体动作识别综述[J].计算机学报,2013,36(12):2512-2524.Hu Qiong,Qin Lei,Huang Qingming.A Survey on Visual Human Action Recognition[J].Chinese Journal of Computers,2013,36(12):2512-2524.
[4]Carlsson S,Sullivan J.Action Recognition by Shape Matching to Key Frames[C]//European Conference on Computer Vision(ECCV).Copenhagen,Denmark:IEEE,2002:263-270.
[5]Wang Xiaogang,Tieu K,Grimson E.Learning Semantic Scene Models by Trajectory Analysis[C]//Proc of the 9th Europen Conference on Computer Vision.Berlin:Springer,2006:110-123.
[6]Shotton J,Fitzgibbon A,Sharp T,et al.Real-time Human Pose Recognition in Parts from A Single Depth Image[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.USA:Colorado Springs,2011:1297-1304.
[7]Jiang Zhuo-Lin,Lin Zhe,Davis L S.Recognizing human actions by learning and matching shape-motion prototype trees[J].IEEE Transactions on Pattern Analysis and Machine Intelligence(S1062-8828),2012,34(3):533-547.
[8]刘鑫辰,傅慧源,马华东.基于RGB-D摄像头的实时手指跟踪与手势识别[J].计算机学报,2014,41(10):50-61.Liu Xinchen,Fu Huiyuan,Ma Huadong.Real-time Fingertip Tracking and Gesture Recognition Using RGB-DCamera[J].Computer Science,2014,41(10):50-61.
[9]Shi Qinfeng,Wang Li,Cheng Li,et al.Human action segmentation and recognition using discriminative semi-Markov Models[J].International Journal of Computer Vision(S0920-5691),2011,93(1):22-32.
[10]陶丽君,李翠华,张希婧,等.基于Kinect传感器深度信息的动态手势识别[J].厦门大学学报(自然科学版),2013,52(4):493-497.Tao Lijun,Li Cuihua,Zhang Xijing,et al.Dynamic Gesture Recognition Based on Kinect Depth Data[J].Journal of Xiamen University(Natural Science),2013,52(4):493-497.
[11]Datal N,Triggs B.Histograms of oriented grdients for human detection[C]//Proceedings of Conference on Computer Vision and Pattern Recognition,Montbonnot,France:IEEE,2005:886-893.
[12]张汉灵,李红英,周敏.融合多特征和压缩感知的手势识别[J].湖南大学学报(自然科学版),2013,40(3):87-92.Zhang Hanling,Li Hongying,Zhou Min.Hand Posture Recognition Based on Multi-feature and Compressive Sensing[J].Journal of Hunan University(Natural Sciences),2013,40(3):87-92.