A target anti‐occlusion method based on image processing and trajectory prediction in electro‐optical tracking system

Abstract

In the electro-optical tracking system, due to the full occlusion of clouds or buildings, the traditional tracking algorithm fails to work effectively. In the authors’ study, a dual-channel anti-occlusion scheme based on image processing and trajectory prediction is proposed. On the one hand, the image-processing unit plays a significant role in evaluating the occlusion state of the target. An improved image occlusion judgement method based on Bhattacharyya coefficient occlusion judgement is proposed, and the ability to resist background interference and rapidity of the algorithm is enhanced. On the other hand, according to the different occlusion states, the dual operation modes of tracking and prediction are usually set. When the target is completely occluded, the prediction mode based on a robust predictive filtering method is investigated to predict the real-time trajectory of the target more accurately. Image occlusion judgement and trajectory prediction can solve the occlusion problems in tracking control. Otherwise, the feasibility and effectiveness of the proposed method are verified by experiments. The cover image is based on the Research Article A target anti-occlusion method based on image processing and trajectory prediction in electro-optical tracking system by Yao Mao et al., https://doi.org/10.1049/rsn2.12235.