Two-dimensional And Three-dimensional Image Technology Based On The Geometry

The rapid development of information technology, people’s understanding of things is not limited to a point, a plane, or even limited to the three-dimensional space, and we always require knowledge and understanding in the field of higher dimensions. Along with the widespread using of intelligent machines in the development of human society, images has become the main carrier of information, and a large number of image information has brought great difficulties to the manual processing, so computer image processing techniques is increasingly attention has been paid. Human visual attention to the image information processing is necessary, in our today’s real life, a growing demand for the conversion from a two-dimensional image to a three-dimensional structure, such as medical imaging, realistic virtual, robot navigation, terrain exploration and so on.Basics geometric reasoning, the theory of geometric algebra is introduced in detail, and the relationship between two-dimensional, three-dimensional space and the geometric is discussed. The image feature extraction, processing and matching has a very important role in the three-dimensional reconstruction. In this article, Harris, Moravec, and Forstner these three feature extraction algorithms were compared and analyzed, the image’s initial coarse matching of the various transformation model, and the image’s fine match of using of the least squares method are analyzed, the use of Ransac algorithm for image stitching, and detailed description of the SIFT feature matching algorithm are realizedBecause the analysis and understanding of the various images, we can easily find that there is only a small portion what we generally interested in of the entire image. The human visual selective attention mechanism is added to the image processing will reduce a lot of the workload, a detailed discussion of the theory of human visual pathway, known Itti model. The article also details the various theories of the three-dimensional reconstruction, and adds Itti model to the three-dimensional model reconstruction, we find that the efficiency is improved.