Existing video description generation methods extract features and feature combinations in a simpler way.resulting in the model losing some of the important semantic information related to the video description, limiting theaccurate description and understanding of the video content. Analysing the deficiencies, this paper proposes a videodescription generation method based on enhanced global-local feature fusion, Firstly, different feature extractors are usedto extract local and global features for the video clips respectively, and in order to model the relevance of different levelsof features (local and global), feature fusion is performed using a feature fusion enhancement network to enrich thefeature information of the model, In this paper, the bidirectional long and short termn memory network used by thedecoder is followed by a reconstruction network, which reconstructs the video feature sequences obtained by the encodelprocessing, and finally generates the descriptive statements of the video through the long and short term memorynetwork. Experimental results on MSVD and MSR-VTT datasets show that the model proposed in this paper cansignificantly improve the accuracy of the generated descriptive statements.