Research on Robustness Recognition Algorithms in Augmented Reality
Shuman Liu, Shuiying Ge, Haitao Yu · 2016
In order to solve the issues of robustness recognition in ARToolkit system, this paper focuses on two main factors: illumination and motion. Firstly, a self-adaptive illumination algorithm is proposed based on the analysis of previous work. By establishing relationship between illumination and grayscale threshold, the algorithm can dynamically adjust the grayscale threshold, thus effectively improve the robustness recognition in ARToolkit applications. Secondly, the method of optical flow tracking algorithm is proposed, which can maintain the robustness of recognition when the object is moving. Consequently, the ideal experimental results show that algorithms are effective and feasible.