Skip to content

TRE computation for point cloud registration #6

Description

@benoitgourdin

Hello,
Thank you for publishing the lung dataset and the associated code!
I am currently working on registration techniques for 3D medical point clouds and am actively using your dataset, as it provides excellent samples for my research. I would like to establish a baseline to evaluate my performance, with the goal of submitting a paper at MIDL by the end of January. I noticed in your paper that you use the TRE (Target Registration Error) value for evaluation.
I have reviewed your code, particularly the pointPWC scripts, but I couldn’t find the implementation for TRE computation. I attempted to write it from scratch, but I was unable to replicate the initial TRE values for the same pair. For instance, for pair 008, I obtained approximately 11 mm instead of the 17.29 mm stated in your README.
Could you please share the TRE computation code with me? Alternatively, is it possible that I overlooked some lines in the repository?
Thank you in advance for your assistance!

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions