Stitching Of Large Image In Photogrammetry Research Of Bridge Engineering Monitoring

In the construction of bridge, real-time monitoring is an important part of ensuring safety and engineering quality. The traditional monitoring method, requires more step and personnel, while measurement accuracy is influenced by artificial factors. With rapid development of computer technology and image processing technology, photogrammetric technology is adopted in bridge deformation monitoring. But for large bridge, it’s difficult to obtain high accuracy of panoramic images. Based on SIFT, image splicing technology and high precision image synthesis problem, the technology is applied to bridge monitoring of. Main work and conclusions are as follows:(1) Applied photogrammetric method to bridge geometry monitoring process, solve the difficulties in traditional measuring method.(2) In view of the image matching technology, proposed the image matching method based on SIFT. The method using the feature points between two images the initial matching, then remove the false matching using the algorithm of random sampling uniformity, and matching the image rotation to benchmark image coordinates, then solve the problem of false match again using random sampling algorithm to remove consistency, the image is realized by using the method of getting fade out.(3) Based on SIFT image splicing method, the image matching and synthetic procedures is improved, and validate the splicing procedures in model test; through the analysis of the actual monitoring and the geometric linear in bridge engineering, and proved the feasibility of the proposed method.