It relates the disordered images and allows the outcomes to be
geo-referenced in a coordinate system, transferring them to a digital elevation
model or ortho-photo. The processing can be developed in few steps, where the
parameters may be adjusted to affect the outcome. Firstly, from the camera, the
images with the correct overlapping, must be imported to the computer. The
interior orientation, or the focal length and image sensor size, is defined
from the Exif meta-data. The image collection process follows by aligning the
images, building the geometry, and if needed, the texture should be added.
Using PhotoScan, geo-referencing can be achieved in two manners,
1. Direct geo-referencing. By
time-stamped GPS data, which is recorded during the flight. The camera and GPS
time synchronize automatically. The imaging exposure position is integrated
into the Exif data as geographical coordinates in the WGS84 format. The
software gathers all data and adjusts it to the exterior orientation of the
images. As a result, a point cloud is carried to the coordinate system.
2. Indirect geo-referencing.
Referencing points are used for this surveying method, either displayed objects
over the ground of the desired area prior to the flight or existing fixed
objects, such as road marks, surveyed using tachymetry or differential GPS.
During the data processing, the reference points must be manually identified in
the software model. The object coordinates are referred to the model, thus, the
full model is geo-referenced by use of a spatial transformation.
For a quick outcome, a
direct geo-referencing method is recommendable. Whereas, if higher accuracy is
needed, the indirect geo-referencing method is more suitable. Current data
processing software may perform a referencing job of 100 images gathered with a
UAV in a less than 3 hr. flight