A FEASIBILITY ANALYSIS OF MOBILE PHONE CAMERA IMAGES FOR 3D INFORMATION

Abstract

As a result of the highly modernized technology, mobile phone cameras have developed tremendously in today’s era. Due to the low cost and considering the various benefits of using them, it is important to investigate the feasibility of using them in digital close-range photogrammetry. This study evaluated the accuracy of mobile phone cameras in comparison with professional cameras in 3D photogrammetric applications. It also attempted to determine how optimal parameters such as distance from the camera to the object, photo overlap, size and shape of the test area influence in 2D and 3D accuracy. Complete workflow consists of four main steps; preparation, data collection, 3D model generation and accuracy analysis. Six different mobile phone cameras and professional camera were used for data acquisition in three study areas. Therefore, 63 3D models were generated under a total of 9 cases. In accuracy analysis, 3D coordinates of generated 3D models using Agisoft were compared with reference points obtained by sokkia fx total station. The final RMSE i.e. (4.238mm-8.448mm) values shows promising results and demonstrated that smartphone cameras can be easily used as low-cost tool in several applications of digital close-range photogrammetry. Further, among the mobile phone cameras used during this study, the iPhone XS max had the highest accuracy i.e. RMSE value was 4.238mm. Also, the obtained accuracy improved rapidly with increasing the overlap percentage between image i.e. higher accuracy was obtained when 80% overlap was used and decreases as distance between camera and object increases. Further, the shape of the object does not affect the accuracy while the size of the object does affect the accuracy. The finding here were limited to small areas, thus, it is recommended to conduct further work for large areas considering different applications.