feat: improve food scanning UX for low-confidence and non-food images

[GAchuzia]

Description

• Add a confidence threshold to the ML prediction so low-confidence predictions (e.g. unclear or non‑food images, lower than 80% confidence score) return an unsuccessful notification banner to the user.
• Align the low-confidence banner styling with the existing buttons so the UI feels consistent.

Note: There will still be edge cases where the current model mistakes non-food items as food with high confidence. The purpose of this PR is to filter out all the low confidence scores.

Fixes #110

Type of change

Please delete options that are not relevant.

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix/feature that would cause existing functionality to not work as expected)
• This change requires a documentation update

How Has This Been Tested?

Please describe the tests that you ran to verify your changes.

• Unit tests
• Manual tests
• Linted/formatted

Scan an obviously non‑food photo (e.g. object or person) and verify:
- The backend returns success: false, reason: "low_confidence", and a message.
- The scan screen shows the red “No food detected” banner and does NOT show a food details table or warnings.

The idea is anything

Checklist

• My code follows the style guidelines of this project
• I have performed a self-review of my code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes

[Indecisive613]

Looks good. It doesn't rule out all edge cases - that would require an entirely new model, but at least we can rule out some very low confidence classifications!