The road network has a complex topological structure, and in monitoring blind spots, this complex topological relationship is difficult to fully present. Therefore, this study proposes an intelligent detection method for blind spots in road network monitoring based on GIS and graph theory algorithms. Firstly, integrate multi-source geographic spatial data and established video surveillance data, and use GIS platform to construct road network layers; Then, the road network layer is abstracted into a graph structure in graph theory, where nodes represent road intersections or keypoints, edges represent road segments and are assigned corresponding weights; Finally, by applying graph theory algorithms such as shortest path analysis, connectivity analysis, and network flow analysis, the road network is analyzed in depth to detect monitoring blind spots and visually display the specific locations of blind spots. Combining the spatial information provided by GIS, graph theory algorithms can calculate the possible paths of targets in the road network, and can also infer and predict based on known path information and topological relationships in the surrounding area. The experimental results show that the over limit distance deviation index of the blind spot detection method studied is closer to the zero line of the unbiased basis, and the Jaccard similarity coefficient remains at the highest level (>0.9), indicating that this method can detect blind spots more accurately.