There are many mechanisms that can be evaluated for solving this problem and this project will provide:
- A literature review of the current state of the art for volumetric estimation
- Rapidly developed Minimum Viable Products (MVP) of the top candidate solutions identified in (1), including evaluation criteria and a testing strategy to understand the cost/complexity and accuracy trade-offs.
- From 2, a demonstratable pre-production implementation for fitting to existing vehicles for real-world evaluation as well as evaluation of the results from this testing
Some potential solutions may include (but are not limited to):
- Multiple fixed video cameras installed in rugged recesses around the tray edge and use OpenCV to develop a DSM of truck payload.
- A moving camera or cameras which are able to get a view of the entire payload surface without placing the camera metres above the load.
- Cameras or other sensors fitted to the loading vehicles to approximate the volume being loaded into the truck rather than fitted to the truck