Text this: Integrating Abstract Meaning Representation to Enhance Transformer-Based Image Captioning