A Low-cost And Large-area Location System Of Multiple Points Indoor

In recent years, the object motion capture technology based on stereoscopic vision and algorithm has played an important role in the field of scientific research, experimental verification, telemedicine, sports training and film production. Due to the optical methods’deficiencies in robustness, the ordinary system is always susceptible to interference in real world application. On the other hand, the accuracy, cost and area of some commercial location systems also limit their technology’s application. In this pager, a new low-cost and real-time multiple points tracking and localization system is introduced while keeping the robustness and instantaneity in a certain degree.The major contributions of our paper are:1. Several infrared LEDs are used to avoid the interference of visible light in order to improve the robustness.2. Some wide-angle lenses of the cameras are incorporated to expand the visual field of cameras.3. The technology of ROI diminishes the image data to be processed in order to promote the instantaneity.4. A real-time multiple points tracking method using reprojection error is used to accelerate the points association among several images captured from different cameras and this theory performs well in the matching test.5. In addition, we use a novel One Point Method with some modifications to decrease the image distortion caused by the wide angle lens, and the above algorithms outperform the others.6. We design a RS232-TTL circuit board to guarantee the communication between the computer and the cameras.In the end, we put these infrared points on the flying robot to locate and navigate it. The experimental result proves that:compared with other location system, our method makes great progress in the aspect of positioning range, cost and instantaneity. And the whole system meets the request of points location indoor in accuracy and instantaneity.Excepting for the above innovative method, the paper also introduce basic theory and research concerning about the optical location and tracking system, which includes the camera model, classic camera calibration, feature extraction, feature matching, stereoscopic triangle method and other3D-reconstruction theory.