【作者】 李莹；

【导师】 顾宏斌；

【作者基本信息】 南京航空航天大学 ， 交通信息工程及控制， 2009， 硕士

【摘要】 本文研究了基于指端标记的视频手指定位技术,在不影响正常操作的触觉及力反馈的同时,准确定位手指空间位置为人机交互提供基础。本文侧重二维手部跟踪与指尖中心定位的研究,系统基于PC平台,选择DirectShow开发包作为系统开发工具。软件采用模块化的设计思想,在DirectShow框架下实现了视频采集、手部跟踪、指尖中心定位等模块。首先对运动跟踪中常用的基于帧差的检测及跟踪,基于减背景的检测及跟踪以及基于肤色的跟踪分别进行比较,最终选取基于目标颜色特征的Camshift算法对手部进行跟踪。为解决当目标运动情况复杂或发生遮挡时跟踪失败的问题,本文提出CamShift和卡尔曼滤波器相结合的算法。同时本文还提出了当跟踪丢失时的搜索框重置以及用户肤色标定等改进方法。经实验证明,改进后的算法具有较强的鲁棒性和实用性。指端中心定位部分,本文以彩色标记区分不同的手指,既降低了识别难度又避免了复杂的特征匹配过程,提高了整体定位速度。算法摒弃受光照影响大的RGB颜色空间,在YIQ、HIS两个空间做对比实验,最终选取具有较强可靠性的YIQ颜色空间进行指端定位。更多还原

【Abstract】 This paper proposes a new finger tracking method with video based on colored markers on fingertips.This method will not affect the feeling and force feedback while tracking the fingers. It is hoped to accomplish the tracking function instead of data glove.The architecture development is based on Direct X standard.This system implements the functions of video acquiring,hand detecting and tracking,centre fingertips tracking,which finally realizes an intelligent system with video finger tracking research based on colored markers on fingertips.The experimental results of frame difference, object’s color character and background subtraction are compared as well as their advantages and disadvantages are analyzed.On the hand tracking research,the CamShift algorithm which means Continuous Adaptive Mean-Shift is mainly discussed. While Camshift is a tracking method based on the object’s color character,and it is widely used for the parameter-less and fast calculation.CamShift algorithm works well in the simple background,but not in the complicated background. So the CamShift algorithm combined with Kalman filter is proposed in my paper. Kalman filter is used to forecast possible position of target, then Camshift search the real position near the possible position. The algorithm has good effect to moving target in the complicated background,and can deal well with simple occultation.Then search-window re-assignment and auto-CamShift are proposed.The results proved that the algorithm is robust and practical.In the fingers tracking,the method of dividing different fingers with colored markers on fingertips is proposed. Besides the reduction of identifying difficulty, it passes over the complicated process of character matching. Different color space,for example YIQ,HIS,is experimented. And YIQ was chose for good result.更多还原