Research On The Key Technology Of Augmented Reality Based On Sensor

With the rapid development of computer technology, Augmented Reality(AR) technology has been more and more popular. As its name said, Augmented Reality adds computer generated virtual object, scene or related prompt information to the real world, in order to augment reality environment. It can blend real scene and virtual object to achieve the consistency and real-time interaction of virtual and real world. The key technology of AR mainly includes: tracking technology, interactive technology and virtual-reality fusion technology etc.. In order to increase the user’s experience in AR system, this paper focus on the research of hybrid tracking based on magnetic and Kinect, virtual-reality fusion based on depth information and interaction based on hand gesture recognition. The main work and innovation of this paper are as follows:This paper proposes an adaptive Kalman algorithm based on fuzzy logic. Traditional single tracking method limits accuracy and robust of the tracking process in Augmented Reality. This paper focus on these questions proposes a hybrid algorithm based on fuzzy logic and self-adaptive Kalman filter, which fuses magnetic data and Kinect depth data to complete accurate real-time hand tracking. Experimental results show this algorithm can effectively use adaptive variance and tracking quality function to improve the robustness of the hybrid real-time tracking and accuracy of the motion estimation.This paper put forward a virtual-reality occlusion algorithm based on the depth information and RGB information. The depth information of the occlusion method is difficult to obtain enough dense depth map in real time, and resulting the occluding edge accuracy is not high in the extraction. Aiming at this problem this paper proposed an algorithm which mix the depth information and color information to get high edge extraction precision. First of all, use the depth map of Kinect to extract a relatively coarse edge; secondly, registration of the color image and depth image to make the depth information corresponded to the RGB image; finally, combining rough contour edge and RGB information, judge pixel belonging, obtain the occluding edge and high precision. Experiments show that, the algorithm can effectively increase the occlusion edge extraction accuracy, gives users a better visual experience.This paper realizes an hand gesture recognition and interaction system based on magnetic tracking and Kinect. Interaction technology usually uses magnetic tracking or Kinect single device. But magnetic device alone can’t complete static hand gesture recognition, it usually used data glove to complete, and dynamic hand gesture recognition of Kinect need to complete complex frame calculation and spatial information conversion, more of that it is also not accurate enough. This paper, aiming at above problems, put forward the magnetic tracking and Kinect hybrid identification to interact. The specific is to use the magnetic and Hidden Markov Model to complete dynamic hand gesture recognition and use Kinect to complete static hand gesture recognition to achieve real-time and better recognition accurate rate. By using the superiority of magnetic and Kinect hybrid recognition to achieve better real immersion and accurate hand operation recognition system.This paper around the magnetic tracking device and Kinect equipment carry out the research of magnetic and Kinect hybrid tracking technology, virtual-reality fusion based on depth information and interaction based on hand gesture recognition. A series of experiments demonstrate the effectiveness and the innovation of the proposed method. In addition, by the mix-used of magnetic and Kinect in AR technology can effectively enhance the robustness and accuracy of the AR registration, which also can greatly promote the development of AR technology.