优质期刊论文发表网提供专业的论文发表、论文写作以及期刊推广服务!QQ:820771224

电话:18796337551
当前位置:优质期刊论文发表网信息系统 → 文章正文

基于Leap Motion控制器的手势交互在应用场景中的研究

作者:优质期刊论文发表网  来源:www.yzqkw.com  发布时间:2019/10/10 9:29:29  

摘要:随着信息技术的持续不断发展,计算机已进入了千家万户,无论是学习还是娱乐,交互是人与计算机唯一的沟通方式。传统的计算机的人机交互方式有很多,例如键盘交互,鼠标交互及显示器交互等,其强制适应性的特点已经不能满足智能及可穿戴设备的需求。人的手势作为人类目前使用最频繁的方式之一,表达方便且直接。本文研究的是基于Leap Motion控制器在交互场景中的应用,在实验过程中结合计算机视觉以及图像处理等相关技术,在采集过程对图像的分割算法进行改进,对单指和双指特征向量进行提取,对动态的手势轨迹进行轨迹捕捉和预测,最终实现了人机交互场景中通过Leap Motion控制器对机械臂进行准确实时的控制,主要工作如下:

1.手势分割图像的研究和改进:本文首先简要介绍并分析了目前国内外基于手势的人机交互的研究现状,之后本文将背景差分法进行了改进优化,提出了将收集的手势数据进行随机性的插入和删除的思想,通过实验验证了该方法可以减少背景图像的干扰,可以有效的在Leap Motion控制器原有的手势识别基础上提高对手势识别的准确性。

2.动态手势的轨迹识别的研究:本文对动态手势的轨迹研究是基于模板匹配的思想,针对比较常用的动态手势,提出了基于Leap Motion控制器的动态手势轨迹的识别方法,经大量实验证明,本文对动态手势轨迹的识别效果比较良好,结果表明该方法的在对近距离的范围内动态手势轨迹准确识别率在96%以上。

3.基于Leap Motion控制器的人机交互场景的实现:最后将之前的研究成果应用到手势交互控制机械臂的控制系统里,引入MPU6050传感器以及Processing对机械臂初始位置以及机械臂在运行过程中产生的位移偏差进行纠正,最终成功搭建了基于Leap Motion控制器的人机交互系统。

With the continuous development ofinformation technology, computers have entered millions of households. Whetherit is learning or entertainment, interaction is the only way for people tocommunicate with computers. There are many ways of human-computer interactionin traditional computers, such as keyboard interaction, mouse interaction anddisplay interaction. Its mandatory adaptability can not meet the needs ofintelligent and wearable devices. As one of the most frequently used ways,human gestures are convenient and direct to express. This paper is based on theapplication of Leap Motion controller in interactive scene. In the experimentalprocess, combining computer vision and image processing technology, the imagesegmentation algorithm is improved in the acquisition process. The featurevectors of single finger and double finger are extracted, and the trajectory ofdynamic gesture is captured and predicted. Finally, the human-computerinteraction scene is realized by Leap. Motion controller controls themanipulator accurately and real-time. The main work is as follows:

1. Research and improvement of gesturesegmentation image: Firstly, the research status of gesture-basedhuman-computer interaction at home and abroad is briefly introduced andanalyzed. Then, the background difference method is improved and optimized, andthe idea of randomly inserting and deleting collected gesture data is putforward. Experiments show that this method can reduce the interference ofbackground image and is effective. Based on the original gesture recognition ofLeap Motion controller, the accuracy of gesture recognition is improved.

2. Research on trajectory recognition ofdynamic gesture: The trajectory study of dynamic gesture in this paper is basedon the idea of template matching. Aiming at the commonly used dynamic gesture,a dynamic gesture trajectory recognition method based on Leap Motion controlleris proposed. A large number of experiments prove that the trajectoryrecognition effect of dynamic gesture in this paper is good, and the resultsshow that the method has a good performance in the near range. The accuraterecognition rate of internal dynamic gesture trajectory is more than 96%.

3. Realization of human-computerinteraction scene based on Leap Motion controller: Finally, the previousresearch results are applied to the control system of gesture interactivecontrol manipulator. MPU6050 sensor and Processing are introduced to correctthe initial position of the manipulator and the displacement deviation of themanipulator during its operation. Finally, a human-computer interaction systembased on Leap Motion controller is successfully built.

关键词:人机交互;机械臂;LeapMotion;背景差分法;模板匹配

Human-computer interaction; Manipulator;Leap Motion; Background subtraction; Template matching

联系方式

客服QQ 820771224
客服热线18796337551
网站地址 www.yzqkw.com
郑重承诺 高效,快速,包发表!
优质期刊论文发表网真诚欢迎新老客户的光临与惠顾!