Improve room differentiation and floor plan accuracy

[TangoEcho]

Problem

The floor plan is still not properly differentiating between rooms - both Bedroom and Bathroom are showing as overlapping rectangular areas instead of distinct room boundaries.

Current Issues:

1. Room overlap: Both rooms show same dimensions (7.19x4.00) suggesting they're using the same floor boundary
2. Poor segmentation: Rooms aren't properly separated despite furniture clustering
3. Missing doorway usage: Not leveraging RoomPlan's door/window/opening data for better room separation

Logs Show:

Room: Bedroom at (3.21, -0.93) size: 7.1938715x0.0x4.004379
Room: Bathroom at (4.28, -2.41) size: 7.1938715x0.0x4.004379

Both rooms have identical dimensions, indicating they're using the entire floor boundary.

Potential Solutions:

1. Use Doorway/Opening Data: Leverage capturedRoom.doors, capturedRoom.openings to create room separations
2. Improve Clustering: Better furniture-based room segmentation with actual boundary calculation
3. Wall-based Segmentation: Use wall data to create natural room divisions
4. Polygon Intersection: Calculate actual room boundaries instead of furniture-padded rectangles

Next Steps:

• Investigate using RoomPlan's opening/door data for room separation
• Consider wall-based room boundary calculation
• Improve furniture clustering algorithm to create distinct room areas
• Test with more complex room layouts

Priority: High - Core functionality for WiFi analysis accuracy.