As the scale of water conservancy projects continues to expand, the amount and complexity of analytical data have also correspondingly increased. Currently, As the scale of water conservancy projects continues to expand, the amount and complexity of analytical data have also correspondingly increased. Currently, It is difficult to realize project management decision support based on a single data source, and most manual analysis methods not only have high labor costs, but also are prone to the risk of misjudgment and misjudgement, resulting in huge property losses. Based on this problem, this paper proposes a visualization and analysis method for unmanned pumping stations on a dynamic platform based on data fusion technology. First, the method uses the transfer learning method to make ResNet18 model obtain generalization ability. Secondly, the method uses the ResNet18 model to extract image features, and outputs fixed length sequence data as the input of long short-term memory (LSTM) model, Finally, the method uses the LSTM model outputs the classification results. The experimental results show that the algorithm model can achieve an accuracy of 99.032%, effectively recognize and classify pump station images, and reduce the occurrence of pump station accidents.