Deformation monitoring for historical site using close-range photogrammetry technique

With the development of remote sensing technology, using optical sensors to record and preserve heritage sites have been widely used and well recognized at an international level. Different from the traditional recording method using hand-painting or plane table to do measurement, using non-invasive optical sensors like digital camera or laser/radar scanner can provide more realistic, accurate and precisely results. Digital recording documents are much easier to preserve and can be utilized in many other purposes like cross-comparison, monitoring of shape and colors, multimedia museum exhibitions, visualization, reconstruction, virtual/augmented reality applications and so on. LiDAR point cloud has real scale and high precision, but its color would be easily affected by weather, levels of brightness, and calibration of the mounted camera. The equipment is also heavy and expensive. The accuracy of image-based point cloud using digital camera is highly depended on the resolution of the images. The texture of 3D model using photogrammetry is more realistic. Digital cameras are also portable. On account of portability, the digital camera is chosen to record the heritage site. This paper conducts a survey in the test field to simulate deformation using terrestrial LiDAR scanner, unmanned aerial vehicle (UAV) and handheld camera to observe movement of measuring points. Aerial images taken by UAV are georeferenced and the image quality is stable but the resolution is not higher than the handheld camera. Terrestrial images taken by handheld camera are easily affected by the photographer (e.g. blurry image because of vibration) and was needed to be given a real scale. Considering all pros and cons mentioned above, the LiDAR point cloud was used to check the accuracy of image-based point cloud by handheld camera and UAV. Through previous experiments, the testing method is feasible. The accuracy of image-based 3D models can achieve centimeter level precision. The high precision measuring results were used to conduct deformation monitoring and displacement observation. Then the accuracy of the 3D models will be analyzed using different kinds of remote sensing sensor data and each method' limitations will be reviewed as well. It can be expected that this can be used to apply deformation monitoring of large scale archaeological remains in the future.