Processing Phantom 3 Drone Data

Introduction
The goal of this lab was to become familiar with unmanned aerial vehicle (UAV) data. UAVs are becoming more applicable in many different scientific fields, therefor it is an important to become acquainted with this technology. The data used in this lab was collected by Dr. Joe Hupy (University of Wisconsin - Eau Claire) using a Phantom 3 drone.

Methods
The Phantom 3 drone was flown at a height of 60 meters. The data was processed using Pix4D software. Once creating the file in the software, the images were added to the file, including the metadata associated with the images. Next, the software assumed that the camera used on the UAV was a global shutter, which was wrong, so it was changed to a Linear Rolling Shutter. The software gathered the coordinate system for the system from the metadata. Once the settings were correct the program showed the overall layout of the flight. At this point initial processing could begin. Initial processing lasted a few minutes. Once it finished, the point cloud and mesh along with the digital surface model, orthomosaic image and index were generated. These last two steps lasted about 15 minutes. Data quality reports were created after each of the 3 processing steps were finished. The final product was two maps created using the processed data in ArcMap.

Results
Two maps were produced using the UAV data; a DSM (figure 1) and a mosaic image (figure 2).

Figure 1. A DSM model was produced using the UAV data. One meter contours were added to accent the changes in elevation.

Figure 2. A mosaic image was created to display the effectiveness of the Pix4D software in generating mosaic images.

Reaction

Pix4D was very user-friendly, meaning that it was not a complex software to use. It generated images that can be used for a variety of purposes. Some of the defaults were too presumptive though, and one must be careful to check all the defaults in the program.