| Virtual Reality technology is the forefront of contemporary information science research fields. Virtual Reality comprehensive application of computer graphics, computer vision, and many other areas of the key technologies in the computer to create a virtual environment to users who feel the true feelings. Virtual reality technology has been present in the military exercises, industrial, medical and health, environmental protection, entertainment, and other areas were widely used. Virtual reality technology includes simulation environment, and perception, natural skills and sensing equipment. Therefore, the core issue is one of the complex real-time generation highly realistic three-dimensional views of the virtual environment. The traditional method of the complexity of modeling, rendering speed and the quality of drawing has become a real-time images generated by impact of the bottleneck. So how to reconstruction the three-dimensional model of objects and scenes will have important significance. Image-based 3D reconstruction is to recover the 3D models of objects and scenes using image sequences. With the traditional use of 3D modeling software or scanner by three-dimensional model of the method, image-based 3D reconstruction based on the method of low-cost, realistic strong and automated processes, which has broad application prospects.In this paper, three-dimensional reconstruction techniques in-depth study, the purpose is to explore a real-time, efficient and convenient sampling of the image-based modeling method to contribute to image-based modeling technology in virtual reality on the actual application, Based on the image for the development of three-dimensional modeling software to provide a prototype system. The main task of this paper is as follows:1. Harris corner of the extraction algorithm and Canny edge detection algorithm to improve, and this combination of the two presents a new algorithm for the application of the automatic image-based modeling algorithm for extracting linear image.2. Presents a wide baseline image layered stereo matching algorithm, and the successful application of image-based 3D reconstruction.3. Based on the development of a 3D image reconstruction software prototype. The software used by the layered stereo matching algorithms and reconstruction based on the results of parameters to gradually amended within the parameters of the camera calibration method. Linear restoration of the 3D geometry of the scene, achieved a 3D reconstruction. |
