Because of the low learning efficiency and easy missing into the local minimum missing, when we use the BP neural network for the motive target tracking, it will influence the accuracy of thetracking algorithm.In order to improve tracking accuracy of BP neural network model, a trackingmodel was presented based on combined Adaboost algorithm and BP neural network.The Adaboost al-gorithm multiple BP neural network of weak classifier formed a strong classifier tracking model.Theefficiency of the proposed tracking model was proved by tracking of moving objects in the video.Thepractical application proves that the BP_ Adaboost algorithm can realize the accurate tracking of mov-ing objects and greatly improve the robustness of tracking algorithm.