[RESEARCH] 3D pose extraction from 2D-image-pixel and distance

[jorgenfj]

Description of task

For TACC we are considering using the depth data from the DVL to extract depth from the 2D-image when following the pipeline. Research the variance in 3D-pose extracted based on varying depth data.

Recommended Reading

• Read up on Camera Matrix

Suggested Workflow

• Setup a ros node that uses an image and corresponding camera matrix, with a constant depth to extract 3D points from pixels.
• Try various depths/angles and see how that affects the outcome.

Contacts

@jorgenfj