基于肤色模型与BP神经网络的手势识别

详细信息    查看全文 | 推荐本文 |

英文篇名：Gesture recognition based on skin color model and BP neural network 作者：张彩珍 ; 张云霞 ; 赵丹 ; 张晓金 英文作者：ZHANG Caizhen;ZHANG Yunxia;ZHAO Dan;ZHANG Xiaojin;School of Electronic and Information Engineering,Lanzhou Jiaotong University; 关键词：离散余弦变换算法 ; 肤色模型 ; 加速稳健特性(SURF)算法 ; 反向传播(BP)神经网络 ; 手势识别 英文关键词：discrete cosine transform(DCT) algorithm;;skin color model;;speeded up robost feature(SURF)algorithm;;back propagation(BP) neural network;;gesture recognition 中文刊名：CGQJ 英文刊名：Transducer and Microsystem Technologies 机构：兰州交通大学电子与信息工程学院; 出版日期：2019-06-10 出版单位：传感器与微系统 年：2019 期：v.38;No.328 基金：国家自然科学基金资助项目(61366006) 语种：中文; 页：CGQJ201906040 页数：4 CN：06 ISSN：23-1537/TN 分类号：146-149

摘要

针对基于视觉的手势识别率不高,鲁棒性欠佳的问题,提出了一种基于YCb Cr椭圆聚类肤色模型分割手势结合反向传播(BP)神经网络识别的手势识别方法。对采集到的图像序列利用离散余弦变换(DCT)去噪处理和边缘检测,根据人体肤色在YCb Cr空间聚类紧凑的特性提取出手势的形状轮廓,将边缘检测与肤色模型分割结果相与得到分割出的手势,利用加速稳健特性(SURF)算法提取构建手势的特征向量,最后通过BP神经网络对手势图分类和识别。实验结果表明:针对复杂背景下的手势,该算法具有较强的鲁棒性,效率高,识别的准确率可达到96%。
        Aiming at the problem of low gesture recognition rate and poor robustness of vision a gesture recognition method based on YCb Cr skin color model of ellipse clustering segmentation gesture and back propagation( BP) neural network recognition is proposed. Firstly,using discrete cosine transform( DCT) for delete noising and edge detection on the captured image sequences. Then extract contour profile of gesture according to the human skin color in the YCb Cr space clustering characteristics,obtain the gestures by combining the results of edge detection with the skin color model segmentation. Extracting feature vectors of gestures by using speeded up robost feature( SURF) algorithm,and finally classify and recognize hand gestures through the BP neural network.The experimental results show that for gestures under complex background the algorithm has strong robustness,high efficiency and the accuracy of recognition can reach to 96 %.

引文

[1] YANG J,ZHU C,YUAN J. Real time hand gesture recognition via finger-emphasized multi-scale description[C]∥2017 IEEE Inter national Conference on Multimedia and Expo(ICME),Hong Kong,2017:631-636.
    [2] GILL S K,SHARMA R. Analysis and implementation of simple gesture recognition algorithms using MATLAB[C]∥2017 the Fourth Inter national Conference on Image Information Processing(ICIIP),Shimla,2017:1-4.
    [3]谭台哲,韩亚伟,邵阳.基于RGB-D图像的手势识别方法[J].计算机工程与设计,2018(2):511-515.
    [4]王兵,董洪伟,张明敏,等.基于Kinect的动态手势识别[J].传感器与微系统,2018,37(2):143-146.
    [5]王海鹏,龚岩,刘武,等.一种时空多尺度适应的手势识别方法研究[J].计算机科学,2017,44(12):287-291.
    [6]杨红玲,宣士斌,梁竣程,等.基于肤色检测与卷积神经网络的手势识别[J].微型机与应用,2017,36(22):58-61.
    [7]张墨逸,张秋余,芦俊池,等.移动单目相机下基于单应的动态手势轨迹识别[J].华中科技大学学报:自然科学版,2016,44(5):7-12.
    [8]刘振宇,崔逊田,李晓辉.基于DCT压缩域的图像边缘检测算法[J].计算机技术与发展,2008(4):111-113,117.
    [9]薛俊韬,纵蕴瑞,杨正瓴.基于改进的YCbCr空间及多特征融合的手势识别[J].计算机应用与软件,2016,33(1):151-155.
    [10] ROSALINA L,YUSNITA N,HADISUKMANA R B,et al. Implementation of real-time static hand gesture recognition using artificial neural network[C]∥2017 the 4th International Conference on Computer Applications and Information Processing Technology(CAIPT),Kuta Bali,2017:1-6.
    [11] ZHANG R,MING Y,SUN J. Hand gesture recognition with SURF-BOF based on Gray threshold segmentation[C]∥2016IEEE the 13th International Conference on Signal Processing(ICSP),Chengdu,2016:118-122.
    [12]王龙,刘辉,王彬,等.结合肤色模型和卷积神经网络的手势识别方法[J].计算机工程与应用,2017,53(6):209-214.
    [13]贾维闯,宫进,吴雄华.基于加速度的BP神经网络手势识别设计[J].电子技术与软件工程,2016(21):94.
    [14]王景芳,施霖.基于人工神经网络对s EMG信号的手势动作识别[J].传感器与微系统,2017,36(6):63-65,69.
    [15] BADI H,HAMZA A,HASAN S. New method for optimization of static hand gesture recognition[C]∥2017 Intelligent Systems Conference(Intelli Sys),London,2017:542-544.