COMPUTER VISION

 

Tech 4 Tracing is supporting the development of computer vision algorithms for the automated detection of identification of explosive ordnance objects. This project, conducted in partnership with VFRAME, is based on 3D photogrammetry and the creation of photo-realistic 3D models to generate synthetic data to train object detection algorithms for objects in a range of conditions and landscapes. 

During the winter of 2021–22 VFRAME and T4T partnered to develop a 3D model of the 9N235 submunition that has frequently appeared in the war in Ukraine. This model was the basis for the creation of synthetic training data for object detection algorithms and high-accuracy 3D printed replicas for benchmarking purposes.

For more information, and to download high resolution images, visit https://vframe.io/about/collaborations/t4t.

3D photogrammetry 9N235/210
Photogrammetry scan of the 9N235 submunition. Image credit: Adam Harvey / VFRAME.io. Image license: CC-BY-NC 4.0.
Sample annotated 9N235/210 submunition. Image credit: Adam Harvey / VFRAME.io. License: CC-BY-NC 4.0.
Initial mission to capture the 9N235 and AO-2.5RT submunitions (right) and cartridge headstamps (left). Conducted in partnership with the Swiss EOD Centre. All munitions pictured are inert and free from explosives. Image license and credit: © Adam Harvey / VFRAME.io.