On my way to work one day I noticed a water tower being built alongside the highway. Since the construction site was not too far from where I live I thought it would be interesting to design a drone mission using Litchi and fly that same mission every day or so as the water tower was being built. This movie is the result of video captured from eight flights over the 100 days of construction of the water tower.
What I didn’t realize when I started this project was how difficult it was going to be to blend the segments taken on different days together. Even though I was flying the same mission and taking off from the same location every flight, the resulting video segments all differed so that I could not simply find the same position in two video segments and the overlay one segment top of the other. Instead, after identifying the same position in two clips captured on different days, I had to offset the X/Y position of one video relative to the one that came before it. This had to be done with each segment I added. Once all of the individual video segments were aligned, it was necessary to increase the scale of the entire montage so that the final video frame was completely filled by the individual segments. The vertical difference of the very last video segment was so great that I had to add a black border at the bottom to fill in the gap resulting from the positioning of that clip. Has anyone else experienced this when trying to do something similar?
I am still happy with the end result. However, if I had to do it over again I would have:
What I didn’t realize when I started this project was how difficult it was going to be to blend the segments taken on different days together. Even though I was flying the same mission and taking off from the same location every flight, the resulting video segments all differed so that I could not simply find the same position in two video segments and the overlay one segment top of the other. Instead, after identifying the same position in two clips captured on different days, I had to offset the X/Y position of one video relative to the one that came before it. This had to be done with each segment I added. Once all of the individual video segments were aligned, it was necessary to increase the scale of the entire montage so that the final video frame was completely filled by the individual segments. The vertical difference of the very last video segment was so great that I had to add a black border at the bottom to fill in the gap resulting from the positioning of that clip. Has anyone else experienced this when trying to do something similar?
I am still happy with the end result. However, if I had to do it over again I would have:
- Used the drone to precisely identify the center of the water tower prior to the very first flight. I could not identify the center using a map because the tower was not yet shown on any map (obviously).
- Used Google Earth Pro to create a “perfectly” circular mission revolving around the tower. Litchi does not provide a way to do this in their Mission Hub.
- Use custom-written software to convert the Google Earth circle into a “perfect” spiral before importing into Litchi. That would have given me a more consistent view of the water tower.