Research and simulation of path estimation based on extended kalman filtering algorithm in video image

摘要

Human beings can use artificial intelligence technology to model, predict and track moving targets. At present, visual target tracking is widely used in military and civil fields, such as visual navigation and positioning, radar target detection, air traffic control, robot visual navigation, video intelligent monitoring and automation, etc. Therefore, visual target tracking has not only theoretical research value, but also practical engineering value. The Kalman filter algorithm is improved to make the tracking effect of moving target better. Firstly, the time difference observation equation and signal arrival direction observation equation are used to locate the moving target, and its mathematical relationship is simplified to generate a pseudo linear model; Finally, the extended Kalman filter algorithm is applied to complete target tracking. Simulation results show that the extended Kalman filter algorithm has better tracking accuracy and stability than the traditional algorithm.