Dr Mingming Liu’s Research Piece on Saliency Detection for Underwater Moving Object
This paper presents a saliency detection framework for monitoring underwater moving objects. The framework combines a tracking algorithm and trajectory analysis to identify objects from background interference. It also includes an analysis model to detect the saliency of local patch sequences and remap it to the global saliency map.
Additionally, the paper describes an underwater sonar data acquisition system to verify the effectiveness and generalisation ability of the proposed methodology. The authors demonstrate that their approach outperforms fifteen other saliency object detection approaches through extensive experiments conducted on eight sonar videos.
Read the full research paper here: https://www.sciencedirect.com/science/article/pii/S0031320324007945