一种时空多尺度适应的手势识别方法研究
发布时间：2018-07-17 05:18
【摘要】：手势交互作为一种自然便捷的交互方式,在智能家居和智能交通等领域具有日益广泛的应用前景。由于手势行为发生的速度、空间约束和用户差异的影响,同一语义手势表现出具有不同时间和空间尺度的多形态特征,这给保障手势识别的准确率带来了挑战。提出了一种基于动态时间规整DTW(Dynamic Time Warping)方法的时空多尺度手势识别方法 SDTW(Spatial-Temporal Dynamic Time Warping),该方法通过对空间形态数据进行分箱操作来达到适应一定程度空间尺度变化的能力。因此,SDTW方法不仅具备DTW方法的时间尺度适应性,而且扩展了空间尺度适应性。文中实现了一个基于智能手机加速度传感设备的SDTW手势识别原型系统。实验测试验证了所提方法能够有效提升手势识别的准确率。
[Abstract]:Gesture interaction, as a natural and convenient way of interaction, has an increasingly wide application prospect in intelligent home and intelligent transportation. Due to the influence of the speed of gesture behavior, spatial constraints and user differences, the same semantic gesture shows multi-morphological features with different time and space scales, which brings challenges to ensure the accuracy of gesture recognition. This paper presents a spatio-temporal dynamic time warping (SDTW) method based on dynamic time warping (DTW). The SDTW (Spatial-Temporal dynamic time warping) algorithm is proposed. The SDTW (Spatial temporal dynamic time Warping) algorithm can adapt spatial scale changes to a certain extent by clocking spatial morphological data. Therefore, the SDTW method not only has the time scale adaptability of DTW method, but also extends the spatial scale adaptability. In this paper, a prototype system of SDTW gesture recognition based on smart phone acceleration sensor is implemented. Experimental results show that the proposed method can effectively improve the accuracy of gesture recognition.
【作者单位】： 西北工业大学计算机学院;西北工业大学软件与微电子学院;
【基金】：中兴通讯产学研合作项目 西北工业大学研究生创意创新种子基金(Z2017183,Z2016194);西北工业大学研究生双创项目,CAST-BISEE基金(2015MC1001061) 中央高校基本科研业务费专项(3102015BJ007) 科学技术基金课题 陕西省科技攻关项目(2016GY-100)资助
【分类号】：TP391.41
[Abstract]:Gesture interaction, as a natural and convenient way of interaction, has an increasingly wide application prospect in intelligent home and intelligent transportation. Due to the influence of the speed of gesture behavior, spatial constraints and user differences, the same semantic gesture shows multi-morphological features with different time and space scales, which brings challenges to ensure the accuracy of gesture recognition. This paper presents a spatio-temporal dynamic time warping (SDTW) method based on dynamic time warping (DTW). The SDTW (Spatial-Temporal dynamic time warping) algorithm is proposed. The SDTW (Spatial temporal dynamic time Warping) algorithm can adapt spatial scale changes to a certain extent by clocking spatial morphological data. Therefore, the SDTW method not only has the time scale adaptability of DTW method, but also extends the spatial scale adaptability. In this paper, a prototype system of SDTW gesture recognition based on smart phone acceleration sensor is implemented. Experimental results show that the proposed method can effectively improve the accuracy of gesture recognition.
【作者单位】： 西北工业大学计算机学院;西北工业大学软件与微电子学院;
【基金】：中兴通讯产学研合作项目 西北工业大学研究生创意创新种子基金(Z2017183,Z2016194);西北工业大学研究生双创项目,CAST-BISEE基金(2015MC1001061) 中央高校基本科研业务费专项(3102015BJ007) 科学技术基金课题 陕西省科技攻关项目(2016GY-100)资助
【分类号】：TP391.41
【相似文献】
相关期刊论文 前10条
1 武霞;张崎;许艳旭;;手势识别研究发展现状综述[J];电子科技;2013年06期
2 ;新型手势识别技术可隔着口袋操作手机[J];电脑编程技巧与维护;2014年07期
3 任海兵,祝远新,徐光，
本文编号：2129185
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2129185.html
