Robbyant / Robbyant/lingbot-map

Loop closure drift in circular trajectory reconstruction

Open
#60 14 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
17.1k
Forks
1.9k
PR merge metrics
No merged PRs in 30d

Description

Image

When reconstructing a circular drone flight trajectory (returning to the starting point), the point cloud exhibits duplicate layers and misalignment at the overlap region. The trajectory path does not correctly close the loop — the end point fails to align with the start point, causing visible drift artifacts.
Environment:
Model: lingbot-map-long.pt
Mode: windowed (--window_size 64, --keyframe_interval 2)
Input: 2873 frames from drone video, circular flight path around a building
Expected behavior: The reconstruction should recognize the loop closure and align the overlapping region to produce a consistent point cloud without duplicate structures.
Actual behavior: The point cloud shows layered duplicates and trajectory misalignment at the loop closure region.

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by reproducing the circular-flight case with lingbot-map-long.pt, 2873 frames, --window_size 64, and --keyframe_interval 2, then trace how trajectory closure and overlap alignment are handled. No file or test is named; done means the endpoint aligns with the start and the reconstructed point cloud no longer contains duplicate structures at the overlap.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
computer-vision, robotics
Issue type
Bug
Difficulty
5/5
Estimated time
Over a week
Activity status
Quiet
Clarity
Needs clarification
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.