期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：2005
卷号：XXXVI-5/W8
出版社：Copernicus Publications
摘要：Nowadays hybrid imaging and laser scanning sensors are common for terrestrial surveying purposes. The basic idea is to complement laser scanning, which directly yields 3D information, with digital image capture, which delivers high resolution images for measurement and texturing. The image sensors are internally built in a scanner or externally mounted on top. The advantage of external mounted devices is the flexibility of the used sensors. Depending on the application, cameras with a different resolution or focal length can be used. Such systems are handy, but the calibration parameters have to be determined or controlled for each sensor and every mount. Hybrid systems offer new possibilities such as monoplotting using a combination of images and 3D point clouds or automatic orientation procedures. Furthermore, the fusion of both data sources will generally lead to a higher accuracy and reliability. These facts emphasize the importance of an accurate calibration between the single sensors of a hybrid system. The calibration can be subdivided into three tasks, calibration of the scanning system, the imaging sensor and the estimation of the eccentricity between the sensors. While scanning systems are usually calibrated by the manufacturer, the calibration of the imaging sensor and the estimation of the eccentricity have to be done on the job by the user with provided calibration software. These calibration tasks are discussed in detail in this paper. Based on the fact that the camera is mounted on the scanner, a comparison between single image and multiple image calibration is done regarding to the accuracy. The comparison is realized with a suitable 3D test field, as it is used in close range photogrammetry. The differences of the single image calibration to the multiple image calibration as reference values are given. The single image calibration is about three times worse in inner and outer accuracy. Further mainly for the eccentricity, a three-dimensional test field is created. In a practical test the eccentricity calibration with and without commonly adjusted camera calibration parameters is evaluated. It is emphasized that the eccentricity parameters are highly correlated with the interior orientation parameters. The most reliable result of the eccentricity calibration could be obtained by integrating the interior orientation of the multiple image calibration fixed
