Abstract
Having a good understanding of how to use GCP points as well as the effect that they have on your data is critical to the data collection associated with any UAS mission. GCP's otherwise known as ground control points allow for help allow for data processing results to be as accurate as possible as it allows for your images to correspond to accurate GPS positions using large stationary markers on the ground. Using one GCP point is not going to help your data collection accuracy at so when deciding to use GCP points it is important to use at least a minimum of 3 points which need to be placed evenly around your flight path to make sure at least 2 images are linked to each GCP point.
Methods
For this lab we did collect our own data at Purdue Wildlife. My flight took place at 350 Meters AGL. After collecting the data it was processed two septette times using PIX4d. The first time the data was processed it was processed not using the GCP points and the second time was using the GCP points. As discovered later on my flight path flown did a poor job at collecting data that captured all the GCP points that where placed. This is discussed further below.
Discussion
This lab was one of the more difficult ones for me in terms the quality of the final product generated. A key issue that I believe was the main reason for the poor quality final product was because of the physical data gathering that was done. During the data gathering what was noticed after completion was the fact that our flight path was way too narrow. Why is this important? When thinking about having such a narrow flight path what happened was the camera of the drone failed to capture 4 out of the 5 GCP points and labeling and then reprocessing the 3d map with the updated GCP points was the whole point of the assignment. That being said I do believe it brings in perspective two important points that are always highlighted in class which are the importance of gathering quality data and then the importance checking the data in the field before you go to process so that if you do come across an error you can fix it by flying the mission again right then and there.
Figure 1: My Camera optimization was way off an I was not able to obtain a check mark in that area. I think it has to do with the overall flight and how I was barely able to get any of the GCPS in the flight path which messed up pretty much the whole lab for me as my deliverables are going to be clearly wrong.
Figure 2: Map view of GCP points.
As you can see in figure 2 all 5 of the GCP coordinates are shown in the image and are indicated by the blue crosses. The problem however is the fact that only one of them will be able to be used as the other 4 GCP point reside outside the boundaries of the flight. This is something that next time needs to be addressed before data processing as having a wide enough flight plan to capture all GCP data points is important to process accurate data. In addition, the image above is the map view of the flight path.
Figure 3: Ray Cloud view of GCP points and data.
Figure 3 is the same image as shown in figure 2 but from the ray cloud perspective which gives us a 3d view of the map created. In this image we can better see how narrow the flight path really was which is why the processing of the GCP points may have been unsuccessful. Furthermore, by expanding the flight path of the mission we would have been able to gather more pictures/data which would have allowed us to better incorporate the placed GCP points on the ground.
Comments