docs(face): add pipeline diagram and algorithm comparison table to face recognition tutorial

[akinshaywai]

Summary

The face recognition tutorial is detailed but dives straight into algorithm theory with no visual overview. New users have no quick way to understand the overall system structure or choose between Eigenfaces, Fisherfaces, and LBPH before reading hundreds of lines of theory.

Changes (documentation only, no code touched):

• Added a Pipeline Overview section at the top of the tutorial with an ASCII flow diagram showing all stages: input → face detection → pre-processing → model training/prediction → output
• Added a confidence score note explaining that lower = better match for all three algorithms, and why a rejection threshold is needed
• Added an Algorithm Comparison table summarising illumination robustness, pose tolerance, minimum training images, and whether adding new faces requires retraining — with a beginner recommendation (LBPH)

No existing content was removed or modified.

Test plan

• Build opencv_contrib docs locally and confirm the new sections render correctly
• Verify no cross-references or existing anchors are broken