After the initial success of the results from the southern trolley building I was very excited to see how the point cloud and model for the northern part of the trolley building would turn out. I was very pleased to find that the results were just as positive as the the previous had been.
As you can see, the point cloud for the northern part of the building is very good. It is also important to note that all of the photos for this model were taken by the multi-rotor. As with the southern model, the multi-rotor was able to achieve excellent photo coverage and also maintain a steady and smooth path.
The side view also gives you a good impression of the angle at which the photos were taken.
And finally, we get to see the northern model with its skin on. Keep in mind this had to be done at low settings because of hardware limitations.
For illustrative purposes I animated flybys of the northern and southern trolley building sections. I believe they give a good perspective of the possibilities of this process and with more processing power the results would be much more detailed and contain less visual noise.
The first animation is of the southern part of the trolley building as shown in the additional field work and proccessing section of the blog.
The second video is of the northern section of the trolley building.
This methodology for collecting photogrammetry data of large scale structures continues to prove itself to be a very viable, cost effective, and robust solution for this problem. It has also shown that you can achieve excellent results with equipment built with a budget in mind.
You will notice from the budget estimate the entire system needed to achieve these results costs about as much a notebook computer, with most of the costs being non recurring. This, I feel, strongly fulfills my goal of making this system affordable to most interested parties.
When I began this project I wanted to achieve several specific goals – the system I configured needed to be affordable, reliable, and viable as a meaningful tool. I have to conclude from the data I gathered during the project that my methodology is sound and completely viable as a useful tool for photogrammetry of large scale structures.