Image Analysis And Fast Implementation Of Tunnel Based On Close-Range Photogrammetry

At present, the methods of tunnel deformation monitoring still rely too much on traditional ground instruments measuring methods. Many traditional technical approaches are high labor intensity with lower accuracy, incompletement, and lack of timeliness, so there are some technical limitations to meet the needs of informationization deformation monitoring. With the development of digital camera and the perfection improvement of digital close-range photogrammetry theory, people started to research on engineering application with a combination of digital cameras and digital close-range photogrammetry.Based on the summation of previous studies, this thesis introduces the digital close-range photogrammetry technique into three-dimensional modeling and deformation monitoring of tunnel construction.The author has studied the key technologies in the whole application process, and developed a tunnel experiment system of digital close-range photogrammetry. The main work in this thesis is summarized as follows:(1) According to the fact that the interior elements of the digital camera are always unknown, and the optical lens distortion are always large and instability, the principle of ordinary digital camera is studied and the program of digital cameras rapid field calibration is developed. The results are compared with the existing commercial camera calibration software and its precision can reach the level of commercial software. It shows outstanding advantages in computing speed, adaptation of calibration images, hardware requirements and some other aspects.(2) Given that the particularities of high gray similarity and mass repetitive texture features of tunnel images, the author has studied on the adaptability of the current mainstream image feature points extraction and matching algorithms. Through the practical experiments of tunnel images, different extraction and matching algorithms are compared and evaluated. To solve the problem of high matching error rate in SURF algorithm, the RANSAC algorithm is introduced to SURF algorithm. Tunnel image and satellite image are used to verify the matching effect of combined algorithm respectively, the experimental results show that we can get the same points with high reliability by RANSAC.(3) To achieve the theoretical algorithm in this thesis by program, the 3D tunnel modeling and deformation monitoring experiment system, which is based on the close-range photogrammetry surveying, is developed. The system includes four modules:camera calibration, image feature point extraction and matching, image measurement and analysis, three-dimensional display, and the system is applied to the actual tunnel engineering for testing.