Maritime safety is crucial for advancing ocean activities and holds significant importance for maritime transportation and marine resource development. Detection and tracking technology provides strong monitoring and early warning capabilities for marine safety, improving the safety and efficiency of marine activities. However, the complexity of maritime conditions leads to a lack of accuracy and stability in current detection and tracking technologies. Therefore, a maritime vessel detection and tracking framework based on multivariate information collaboration is proposed to improve the performance of maritime target detection and tracking, providing important references for maritime safety. The proposed method achieved leading evaluations of MOTA 83.7% and IDF1 71.7% in the JMT2022 dataset, and ablation experiments also confirmed the effectiveness of the proposed method.