The paper presents an efficient multi-object tracking (MOT) system for autonomous vehicles, focusing on the integration of LiDAR data and Bayesian methods. The authors introduce a novel measurement model called the Box Data Association Inflate (BDAI) model, which assesses each target's state and trajectory without noise. The BDAI model is complemented by a Box Object Filter method to avoid ambiguous detection responses during data association. The system is designed to optimize computational complexity for continuous object monitoring and tracking. The proposed system is evaluated using the NuScenes dataset and a lab dataset, demonstrating superior performance in terms of tracking accuracy, computational efficiency, and power consumption compared to state-of-the-art methods. The BDAI model achieves 58.09% tracking accuracy and 71% mAP with a pre-processing time of 5 ms, while the Jetson Xavier NX device consumes 49.63% GPU and 9.37% average power, exhibiting a latency of 25.32 ms. The paper also includes a detailed theoretical derivation of the Box Object Filter and the BDAI model, along with experimental results and comparisons with other MOT systems.The paper presents an efficient multi-object tracking (MOT) system for autonomous vehicles, focusing on the integration of LiDAR data and Bayesian methods. The authors introduce a novel measurement model called the Box Data Association Inflate (BDAI) model, which assesses each target's state and trajectory without noise. The BDAI model is complemented by a Box Object Filter method to avoid ambiguous detection responses during data association. The system is designed to optimize computational complexity for continuous object monitoring and tracking. The proposed system is evaluated using the NuScenes dataset and a lab dataset, demonstrating superior performance in terms of tracking accuracy, computational efficiency, and power consumption compared to state-of-the-art methods. The BDAI model achieves 58.09% tracking accuracy and 71% mAP with a pre-processing time of 5 ms, while the Jetson Xavier NX device consumes 49.63% GPU and 9.37% average power, exhibiting a latency of 25.32 ms. The paper also includes a detailed theoretical derivation of the Box Object Filter and the BDAI model, along with experimental results and comparisons with other MOT systems.