Photogrammetry is the use of a two-dimensional medium—photography—to survey and measure physical objects in order to then obtain three-dimensional data-sets or digital models. It is an old technology, dating back to the early days of analog photography. Now that drones are more affordable and photogrammetry software is more sophisticated, the tool is available to a wider public. This digital pointillism aims for accuracy, which is the result of data collection settings and weather conditions: the height set for the flight, the overlap percentage, camera specifications, wind speed, precipitation, and other factors. The decisive factor, however, is battery life: when surveying, Mavic can only be in the air for roughly ten minutes before it shuts down. More fly-time would mean more precision in the model.
In post-production, a software is needed to patch together the images: Pix4D is its name. It highly dislikes reflective surfaces, moving elements, and complex geometries. When photographs contain these elements, the resulting point clouds are imperfect and full of blanks. It is no coincidence, though, that the gaps are mainly found in areas of dense vegetation. Pix4D has no trouble rendering out the human made—roads, buildings, cars. The gaps appear when not enough similarities are found in the photographs’ overlaps and it becomes too difficult to identify distinct elements that allow images to be stitched together. Even though it is a “mistake”—and usually a quite undesirable one—I find it beautiful that trees don’t want to conform to being objects and that they resist becoming individuals—including in photogrammetric terms