VIEMF: Multimodal metaphor detection via visual information enhancement with multimodal fusion