Improve image detection logic

[Strift]

Pull Request

Related issue

Fixes #613

What does this PR do?

• Make image field detection logic recursive to find nested image fields
• Move image detection logic at the hit level - to handle documents with different structures
• Add relevant unit tests

Image fields are only scanned up to the 10th level of depth in the object to avoid impacting performance too much. We can likely set this higher, but I'm not sure it's needed. Your feedback is welcome @CaroFG

PR checklist

Please check if your PR fulfills the following requirements:

• Does this PR fix an existing issue, or have you listed the changes applied in the PR description (and why they are needed)?
• Have you read the contributing guidelines?
• Have you made sure that the title is accurate and descriptive of the changes?

Thank you so much for contributing to Meilisearch!