Research On Laser Pointer Interaction Technology

Laser pointer interaction is a human-computer interaction technology, which combines technologies such as image processing and pattern recognition with a laser pointer. The system has the advantage as follows:no restriction of desktop computing model distance, high mobile flexibility and high pervasively, which is widely used in the areas of large screen interaction, multimedia presentation, robot navigation, virtual reality and smart home. However, problems related to delay, accuracy and misjudging still exist in the traditional laser pointer interaction. Hence, it is a great challenge that how to make the system more robust with better real-time.There are three principal defects of the traditional technology:detection error brought by the high-frequency changes of the background, inaccuracy of coordinate mapping and misjudging of interactive commands resulting from the methods for analysis of laser behavior. This paper proposes an improved laser pointer interaction system to enhance the real-time and robustness. Firstly, a robust dictionary construction and update algorithm based on the sparse model of background subtraction is presented to detect laser spot. In order to control dynamic backgrounds, we determine whether there is a change in the backgrounds; if this is true, the new background can be directly added to the dictionary; otherwise, we run an online cumulative average on the backgrounds to update the dictionary. Secondly, a new coordinate mapping method is used to improve the mapping accuracy. The matching points between computer image and projection image are extracted using SIFT, which are used to train the regularized neural network for obtaining the mapping function. Lastly, the projection area is divided into two sections, control area and label area, and behavior of laser spot in different areas is applied to realize interactive commands.The experiment is done in the environment of large screen interaction and multimedia presentation to test the laser spot detection algorithm. It confirmed the superior performance of the proposed method in terms of detection error and real-time implementation. Subsequently, this paper compares the accuracy of different coordinate mapping algorithms with the camera located differently, and the proposed algorithm shows the best performance. Finally, the laser pointer interaction system is achieved through the plarform of Visual C++ and OpenCV library, which shows good real-time and robustness for accomplishing different interactive commands.