Calibration of nonmetric cameras used in unmaned aeiral vehicle

The use of Unmanned Aerial Vehicles (UAV) has advanced today with the developing technological possibilities. The concepts of time, cost and accuracy are essential in engineering studies. With the UAV, high accuracy is achieved in less time, even in difficult topographic conditions. Integrated cameras are critical equipment due to their effect on accuracy and cost in photogrammetry activities with UAVs. The camera on which the image acquisition is performed is of great importance to obtain high accuracy data from UAVs. The digital cameras used in integration with the UAV are mostly light and low-cost non-metric cameras. Lens deficiencies such as distortion are more common in non-metric cameras. An appropriate calibration method should be determined to reduce the effects of distortion-like lens defects on the final products, and the calibration of the integrated UAV camera should be performed. The parameters obtained by camera calibration (focal length, principal point coordinates, radial and tangential distortions) are a requirement in photogrammetric studies for an integrated UAV camera. It is seen as the most preferred method of self-calibration among camera calibration methods. In most researches, results have been obtained sometimes by establishing its calibration system and sometimes using a ready-made calibration system. In this study, using six calibration test fields and two various UAVs, the software library's self-calibration method and calibration values were tested. Accuracy analyses were performed according to bundle block adjustment with assistance beams using calibration parameters obtained from the calibration test areas created. Subsequently, different results were obtained by using different calibration values with the self-calibration method. The comparisons were made on the 3D model produced from a rough terrain of 215 m wide and 215 m long at different heights in the field. While creating the model, 43 Ground Control Points (GCPs) were established in the area, and the locations of GCPs were determined by the fast static GNSS method. After the image acquisition, most of the GCPs were used for adjustment purposes and the rest for check-point to produce a 3D model. Also, in the evaluation of these images, a self-calibration application was made from the calibration step. The effect of the images obtained as a result of different calibration methods on the model's accuracy was investigated. For both UAVs, the self-calibration way is of acceptable accuracy, according to the laws in use.