Track-before-detect algorithms for underwater targets:An overview
In response to the complexity of the marine environment and the low detectability of underwater targets,underwater target detection and tracking have become urgent problems to be solved in underwater attack and defense operations.Track-before-detection algorithms adopts a trajectory search based on tracking ideas to avoid complex data associations;The use of multi frame data to accumulate decisions and improve the probability of correct detection is playing an increasingly important role in low detectability target detection and tracking.Firstly,the principle of track-before-detection algorithm was explained,and the challenges of low detectability of underwater targets to existing detection methods were analyzed;Secondly,five typical track-before-detection algorithms based on dynamic programming,Hough transform,particle filtering,finite set,and histogram probability multi hypothesis tracking were analyzed and compared,with a focus on their principles,characteristics,applicability,and improvement algorithms;Finally,by exploring new methods for track-before-detection of underwater targets,the robustness,accuracy,and real-time performance of existing algorithms will be improved.Research on the fusion detection and tracking of multiple targets,maneuvering targets,and sensors will be conducted to discuss and prospect underwater target track-before-detection algorithms.
万方数据
维普
