Enhancement of Image Processing Procedure for Multiple UAV Flying Blocks

Processing of a vast number of aerial images from high-resolution digital camera always leads to the use of lengthy hours. This contradicts to the advantage of the use of UAV that could acquire and deliver the data rapidly to highlight the actual condition of the terrain. In previous studies, 1033 images were used covering the size of about seven kilometres square of the study area. During image acquisition the time taken was about three hours while for the image processing it took more than 12 hours to be finished. Image processing was tested to be run in normal computer platform by image processing software that allows for automated bundle adjustment of aerial images considering the interior and exterior orientation of the camera has been resolved. However, due to the huge amount of file size, the output product cannot be displayed correctly after processing stage. Thus, in producing the output model of Digital Orthophoto and Digital Elevation Model of a large area at large scale, many precaution and guideline must comply so that the data could be processed thoroughly until the perfect match of the terrain model could be generated. From this studies, it can be concluded that during image processing stage, not all of the images need to be processed altogether if the overlapping percentage of the adjacent images are already high. Besides, the processing stage must be done by considering the image resolution used which also influencing the size of the output model.