Introduction
This lab has the same premise as the previous lab post as the purpose of this lab was to examine the relationship of different flight parameters and settings and how it affects the data processing of a 3d model using the software Pix4D. In this lab however different mission settings where used in order to collect data. These mission settings will be further discussed below under the Methods section of this post.
Methods
The parameters that we followed for our mission also were to fly at an altitude of 10 meter with a lens angle set to 80 degrees and an overlap of 90%. This high overlap percentage is key to allow for a complete processes image regarding not leaving out any details when the mission was being flown. Prior to the mission being conducted all the different teams debriefed to go over all the missions that would be flown at varying overlaps, lens angles, as well as the pattern that the mission would be flown with. For example, some groups were flying the mission in a grid pattern and some groups were flying the mission using an orbital technique. With several groups flying at the time, it was important to be able to communicate altitudes as well as mission intentions to mitigate any potential accidents. Furthermore, some larger UAS, specifically an M220, were reported in the vicinity of the mission.
Discussion
Figure 1: This is the initial processing of the data in the form of a 3d model. The dots represent ech individual image being processed and will change colors depending on where in the processing stage the program/software is.
Figure 2: I analyzed the quality report and looked for potential eros in the data processing. After looking at both of these side by side I noticed that in the back top corner there is a cluster of red dots signifying there was an error with the data captured. Due to the fact that these data points showing errors are clustered together like this I would assume it has less to do with each individual data point and more with that specific area regarding where the data points were captured. We also see a small cluster of red in the top right middle.
Figure 3: Circled in yellow is a zoomed in image of the clusters of the data points that are having trouble processing. As you can see these areas create a hole if you will in the model as when I try to enable the triangular meshing these areas do not have any matching sides to form a more complete model. Furthermore when looking at this zoomed in image something I notice is how the edge in the top corner seems more rounded than they do anywhere else in the model. The significance of this when thinking about it in both the context of the flight conducted to gather the data and then the actual processing of the data is that maybe potentially during the flight the path of the drone did not take a fine sharp enough turn at this specific point as it may have been the beginning the crosshatch section of the flight.
Figure 4: During this part of the lab is when I ran into some processing difficulties that I was unable to resolve, and if able to resolve would enhance the quality of the 3d model greatly. What I encountered was that I was unable to enable the triangular mesh which would make the model look much less grainy and more realistic. I used different resources to try and help resolve the issue such as PIX4d help documentation but that did me no good.
Comentários