Abstract
Performance of any system can be enhanced by factors such as accuracy and latency. In past six decades, many papers have been published that compares the detection accuracy algorithms and method but very few information is presented about the latency and the methods with which it can be reduced. Dynamic 2d, 3d marker detection through camera has attracted much attention nowadays. This paper presents a fast 2D, 3D marker recognition at least time slot (latency) for augmented reality applications. The proposed algorithm is stimulated by the A forge.net framework and designed against the challenging occlusion and illuminous condition of marker at least delay time. To achieve the sharp edges detection used Otsu thresholding approach. Position of dynamic marker with respect to camera is calibrated through Pose estimation approach. Numerous experiments have been piloted on recognition of 2D, 3D markers. The result analysis computes the performance of the system through enhancing accuracy & latency.
Subject Classification: