Yunnan's minority artifacts, as symbols of national memory and culture, have not been adequately preserved or utilized. This paper aims to construct a knowledge graph for the image description of Yunnan minority artifacts. First, a typical artifact ontology was developed using existing data. Then, traditional BERT-based methods were compared with emerging LLM-based approaches for entity recognition and relation extraction. The experiments show that using BERT combined with Bi-LSTM and CRF for entity recognition and BERT for relation extraction outperforms LLM in this scenario. This study adopts traditional methods to successfully construct a knowledge graph, enhancing the comprehension and interpretation of these cultural elements through precise entity linking.