增强现实运动头部目标跟踪中的误差消除方法

Head Target Tracking Error Elimination Method in Augmented Reality Movement

研究图像目标的跟踪优化定位问题。由于目标运动过程中的方向不定,给实时定位获取目标的位置信息带来了难度。传统的基于SIFT特征点的跟踪定位中,由于人员的运行方向不定,再加上人员运动高度不定,使得特征匹配过程的空间转换的过程计算量很大,与后期的定位形成一个时间滞后,不能满足实时性准确性的要求。提出了一种时间域误差补偿增强现实运动目标跟踪算法。先利用自然特征同步跟踪方法,描述增强现实运动目标跟踪的主要过程,实现目标运动信息与虚拟现实进行融合。引入一种时间域补偿函数,对增强现实运动目标跟踪中的误差进行消除处理,从而完成增强现实运动目标跟踪。实验结果表明,算法进行增强现实运动目标跟踪过程中的误差消除,能够提高增强现实运动目标跟踪的实时性与准确性。

This paper researched the location problem of image target tracking. The target direction is indefinite in motion, which brings the difficulty to obtain the real-time positioning location of the object information. This paper presented an algorithm of time domain error compensation for augmented reality movement target tracking. To use the natural characteristics of synchronous tracking method, we described the main process of augmented reality movement target tracking, realized the target motion information fusion with virtual reality. This method introduces a compensa- tion function in time domain, makes the error elimination processing of augmented reality movement target tracking, thus completes the augmented reality movement target tracking. The experimental results show that the algorithm of error elimination of augmented reality movement target tracking can improve the real-time and accuracy of the aug- mented reality movement target tracking.