added custom data logic

[nico0704]

Simple Custom Dataset Compatibility

The original algorithm was designed with ground-truth poses and specific evaluation in mind, which created challenges for custom data integration (Issue #36, Issue #26, Issue #45). This update introduces the most simple way to handle custom dataset functionality across key parts of the code.

Summary of Changes

• Modifications:
  • gs_icp_slam.py: Adjusted for custom image loading and initial pose settings.
  • mp_Mapper.py: Adapted to handle custom start poses, custom image directories, and adjusted metrics calculations.
  • mp_Tracker.py: Updated to support custom start poses and exclude specific evaluation methods not applicable to custom datasets.
  • traj_utils.py: Added function for loading hardcoded default pose

---

Collaborators: @jojo892, @leon-kratz