Five-point Relative Orientation Based On Forward Intersection And Applications

The accuracy of Videogrammetry System can only be increased from two aspects when the system is under certain resolution. One is to improve the calibration and error correction precision of Videogrammetry System. The other is improve location and matching precision of key points in the image. Therefore, the main contents of this paper include location, matching the key points in high precision and camera calibration. Specific studies include the following aspects:Researched algorithms to extract and subpixel locate artificial marks. The use of artificial markers can not only improve the accuracy and reliability of the measurements, but also can significantly improve the measurement efficiency, automated measurement because artificial markers are easy to layout. This paper focuses on algorithms to extract and subpixel locate code point and verify effective and robust of the proposed algorithm by experiment. Code point which is location in high-precision and match well will be used to calculate the relative orientation elements of the picture pair.Researched algorithms to match feature point in multi-view. Using multi-scale Harris corner detection operator to extract strong feature points which are insensitive to projection and radiation distortion. Feature points are described with SIFT feature extraction method’ description. And finally use the pixel information of feature point and multi-view geometry of the picture pair to complete the match. Verify effective and robust of the proposed algorithm by experiments.Proposed a five-point relative orientation algorithms based on forward intersection. The coplanar equations with forward intersection constraints are set up, whose optimization function is derived, and the least square generalized inverse method is employed to achieve relative orientation parameters with five homologous points. And verify effective and robust of the proposed method by experiments.