Search

In recent years, a range of UAV photogrammetry applications, such as long-range corridor mapping, has increased significantly [4]. However, in photogrammetry linear configuration of the flight is considered to be problematic, due to geometric distortions creating “bowl effect” and a high correlation between focal length and flight altitude. Thus, it is important to investigate how factors like georeferencing method and camera calibration affecting final accuracy. The aim of the article is to determine if the direct georeferencing method and lately adopted in Agisoft Metashape Fourier additional correction model with 96 parameters allow eliminating bowl effect. During the research, two neighboring flights of 5 and 2.5 km railway lines were made. The flights were performed with a fixed-wing UAV equipped with a Sony RX1RII camera and PPK receiver on-board. The area of the research was covered with 68 ground control points which were used for accuracy analyzes. First of all, the results show that the relevant camera distortion model is crucial for eliminating bowl effect. Regardless of the georeferencing method, the use of additional camera calibration corrections significantly increased the accuracy of the adjustment. Secondly, the advantages of using precise camera coordinates were proven. It allowed to decorrelate external and internal orientation parameters and fulfill accuracy requirements even without ground control points. On the other hand, the classical scheme with indirect georeferencing showed high sensitivity on the number and distribution of the ground control points. [ABSTRACT FROM AUTHOR]

It is known that the fundamental photogrammetric and remote sensing problem is the determination with high accuracy of the attitude, the position and the intrinsic geometric characteristics of the sensor. The new aerial photogrammetric systems allow the real-time determination of all parameters of exterior orientation with good enough accuracy and for this reason we can achieve direct georeferencing without needing to determine the coordinates of landmarks in the field, before or after the photogrammetric flight. In our application in this paper we consider known the camera's interior orientation parameters, the camera's exterior orientation parameters (position and attitude) and the boresight misalignment angles between navigation sensor and digital camera system. So, we present some practical results of so called direct georeferencing of digital images, using of direct measurements of the image's exterior orientation parameters by a GNSS/IMU system for DTM obtaining and a fast automatic image's orthorectification. In our application we used the collinearity concept, which is the basis for all direct georeferencing formulas, without using ground control points. We used ground control points only for a good accuracy estimation of the method. As a conclusion is that most of the actual developments of digital optical cameras, or another type of sensor, have integrated the IMUs and the use of GNSS/IMU for direct georeferencing of airborne sensor data is nowadays standard. The accuracy of this rectification result, through a direct georeferencing method, is crucial for overlaying the data with any existing data sets or maps and using them for evaluations in change detection, map updating or other purposes. [ABSTRACT FROM AUTHOR]

The Geometric quality of a photogrammetric product is given mainly by the accuracy of the determination of the sensor orientation. There are several ways how to determine the exterior orientation parameters of aerial frame cameras. In the past, the most frequent option was the aerotriangulation with the bundle block adjustment based on tie and ground control points. At the begining of the 21st century there was a mass commencement exercitation of the on-the-board GNSS and IMU devices and a few new possibilities have emerged. However, the simplest solution comprises direct georeferencing from GNSS/IMU data but many experiments have confirmed that superior results are given by integrated sensor orientation. It means digital aerotriangulation with the GNSS/IMU orientations as the observations. Observations of several ground control points are proposed to eliminate systematic errors and to assure absolute accuracy of the whole block. The GNSS positioning refers to the 3D rectangular earth-centred coordinates in the WGS84 coordinate system. However, the results of the block aerotriangulation are, due to the cartographical demands, mostly required in the national vertical and horizontal reference system. Therefore, GNSS/IMU data has to be transformed to the desirable coordinate systems before the integration to adjustment. Nevertheless, computations in the national coordinate system cause a few problems because the geometric model of block adjustment operate within cartesian coordinate system with the identical scale of each axis. As all the modern national horizontal coordinate systems are conforming with some scale error, the condition of the identical scale is not acomplished. Furthermore, if the gravity-related vertical datum is used, not even the condition of orthogonality is fulfilled. While non-orthogonality has practically no impact, different scale of positioning coordinate system seriously affects altitudes of image projection centers and causese biases in the block adjustment. There are two ways how to avoid this effect. The first one is based on the camera focal length modification and the second one on changing the vertical coordinate system scale. Unfortunatelly, none of these exact methods are commonly not been used in practice. The article looks at modification of altitudes of aerial photograph projection centers before integrated sensor orientation. Numerous outcomes affirming the theory are performed. [ABSTRACT FROM AUTHOR]

The method for spatial relation determination of equipment of an autonomous mapping airship is presented in the paper. The selection of an airship as a carrier was based on our specific requirements such as carrying capacity, long flight time, safety (in case of failure there is no danger of straight fall, airship staying heavier than air therefore it will drop smoothly to the ground) and flight characteristics such as stability of flight in terms of vibrations and possibility to flight at low speed. On the airship a measuring platform is mounted in the damped gravitational gimbal. The platform has modular concept which allows employing different sensors and changes of its amount and position. The only limitation is total weight, which shouldn't exceed 15 kg (carrying capacity of the airship). Currently, equipment consists of tactical grade INS/GPS navigation unit (iMAR iTracer - F200), laser scanner (SICK LD-LRS1000) with conical modification, digital camera in the visible spectrum (Canon 100D, Olympus E-PM1) and professional thermometric camera (FLIR SC645). The aim of the whole project is to create measuring system which will have specific properties suitable for effective mapping of medium-wide areas (units to tens of square kilometers) e.g. [15]. This system should be useful especially for ranges that are too large for conventional surveying with GNSS or total station and too small for using of manned air vehicles from economical and accuracy point of view. The methods for spatial relation determination between pose estimation unit INS/GPS and used sensors are presented in the paper. The methods are based on direct laboratory measuring of coordinate offsets (lever-arms) and in-flight determination of rotation offsets (bore-sights). Bore-sights are determined from comparison between INS/GPS unit and sensors poses, when sensors pose is determined from its observations. Used methods for camera and scanner bore-sight determination are based on the state-of-the-art methods presented in the literature. [ABSTRACT FROM AUTHOR]