A Target Tracking Filtering Method Based on Deep Learning with Denoising Autoencoder
With the rise of deep learning in recent years,deep learning models have also been applied to target tracking and have made significant progress due to their powerful learning ability for complex patterns.However,existing deep learning-based target tracking models use recurrent neural network approach,and this modeling approach does not have enough ability to handle the noise in the highly noisy target observation data and to recover the real trajectory.To address the above problems,this paper pro-poses a de-dry self-encoder-based trajectory tracking model,which adopts a code-decode data processing approach,firstly,a cod-ing network removes the noise from the data and learns the main features of the target motion,and then a decoding network is used to restore the real motion of the target.The results show that the proposed method can effectively reduce the mean square error of tar-get tracking and improve the tracking effect.
万方数据
维普
