Recent research conducted in ideographic languages such as Chinese has underscored the significance of glyph information for Named Entity Recognition (NER) tasks. Nevertheless, the majority of current techniques focus solely on the glyph information of individual Chinese characters, overlooking the contextual relationships among glyph information within sentences and the connections between glyph information across characters in the lexicon. This study proposes Glyph Lexicon Enhanced BERT (GLEBERT), an innovative model that incorporates diverse interaction information of Chinese character glyph into the glyph representation of characters. This approach not only extracts glyph information from individual Chinese characters but also integrates the intrinsic connections between glyph within the lexicon, thereby enriching the representation of Chinese character glyph. Empirical results from four Chinese NER datasets show our model achieves state-of-the-art efficiency for Chinese NER. Furthermore, we conduct a series of experiments to analyze the impact of varying GLEBERT components and parameter configurations.