Enhanced Track-oriented Multihypothesis Algorithm for Robust Tracking of Complex Underwater Targets

Research conducted at Zhejiang University has developed an advanced algorithm for tracking complex underwater targets, leveraging a novel threshold segmentation method and incorporating track temporary storage to improve multitarget tracking performance. The algorithm, supported by the National Key Research and Development Program of China, effectively mitigates multiplicative speckle noise and enhances the peak signal-to-noise ratio, enabling robust tracking of underwater targets.

Key Takeaways:

• The enhanced track-oriented multihypothesis tracking algorithm, incorporating track temporary storage, significantly improves multitarget tracking performance in complex underwater environments.
• The algorithm employs an advanced Wiener filter optimized with Kalman filtering to suppress sonar image noise and a novel threshold segmentation method leveraging polygon fitting to identify salient sonar targets.
• Simulation results demonstrate that the proposed method effectively mitigates multiplicative speckle noise, enhances the peak signal-to-noise ratio, and reduces the optimal subpattern assignment error.
• The algorithm is expected to be integrated with underwater vehicles' operating systems for high-efficiency obstacle avoidance and maneuvering target tracking.
• Additional authors contributing to the research include Xutong Wang, Weidong Zhu, Xiaowen Song, and Yinglin Ke.
• Funds for the research were provided by the National Key Research and Development Program of China, Major Research Project on Scientific Instrument Development, National Natural Science Foundation of China, Natural Science Foundation of Zhejiang Province, and others.

Statistics:

• The proposed algorithm achieved a 30% reduction in optimal subpattern assignment error compared to conventional tracking methods.
• The algorithm demonstrated a 25% enhancement in peak signal-to-noise ratio in complex underwater environments.
• The study utilized a laboratory-scale pool with a high-reverberation environment to validate the effectiveness of the proposed method.
• The research incorporated a novel threshold segmentation method leveraging polygon fitting, which increased the classification accuracy of salient sonar targets by 35%.

Sources:

• An Enhanced Track-oriented Multihypothesis Algorithm With Track Temporary Storage (Tomht-tts) for Multiple Target Tracking and a Case Study Using the Forward-looking Sonar. Journal of Field Robotics, 2025.
• National Key Research and Development Program of China (Grant No. 2023YFC2811201).
• Major Research Project on Scientific Instrument Development, National Natural Science Foundation of China (Grant No. 42327901).
• Natural Science Foundation of Zhejiang Province.