Skip to content

Latest commit

 

History

History
17 lines (13 loc) · 1.19 KB

README.md

File metadata and controls

17 lines (13 loc) · 1.19 KB

Tarp Photogrammetry

Adventures in drone photogrammetry using machine learning and Rust

This code is in support of experimenting with basic photogrammetry methods using a DJI Mini 2. There are two photos, under photos/original, where DJI_0137.JPG is taken at an altitude of 9.9 m and DJI_0139.JPG is taken at an altitude of 20 m. Once these parameters, along with the scaling_factor and tolerance are set in main(), running this code should produce an estimate for the area of the tarp, which has a measured real area of 3.96 m2.

An in-depth explanation of this code can be found here, or in raw form under the git repository at https://github.com/quietlychris/site.

Segmented tarp at 9.9 m height, with an area estimate of 3.84 m2 vs. 3.96 m2 actual

Data is included, and results can be replicated using

$ git clone --depth=1 https://github.com/quietlychris/tarp_photogrammetry.git
$ cd tarp_photogrammetry
$ cargo run --release

Execution time varies on image size, and may exceed several minutes.