feat(plot): add share_axes to show_segmentation for synced viewing

[shcrela]

Description

This PR introduces synchronized zooming and panning to the show_segmentation function by adding an optional share_axes parameter.

When analyzing high-resolution microscopy data, users frequently need to inspect fine details across different views (original image, masks, flows, and outlines). Currently, these subplots have independent coordinate systems. Enabling share_axes links them, allowing for a "zoom-one-zoom-all" experience in interactive Matplotlib backends.

Changes

• New Parameter: Added share_axes=False to show_segmentation.
• Refactor: Reorganized the subplot creation logic into a loop. This reduces code duplication and allows subsequent axes to cleanly reference the first axis (ax0) for sharing.
• Documentation Alignment: Added the missing seg_norm parameter to the function signature to match the existing docstring.
• Preservation: Maintained all original image processing, normalization, and file-saving logic to ensure no breaking changes for existing workflows.

Testing

Verified that:

1. Panning/zooming on the original image updates the masks and flows views when share_axes=True.
2. The function still behaves as expected (independent axes) when share_axes=False.
3. File saving via file_name still outputs correctly.

[mrariden]

Hi @shcrela please share some images of the output before and after to help visualize the improvement

[shcrela]

Actual zooming and panning:

Proposed, synced, zooming and panning: