Underwater Target Detection Method Based on Dynamic Channel Impulse Response Analysis

By constructing a time-varying channel background model and analyzing the dynamic channel impulse response, a refined image is generated, which solves the problem of detection accuracy for low, slow, and stealthy targets in underwater acoustic detection, and achieves high-precision underwater target identification and velocity estimation.

CN121661481BActive Publication Date: 2026-05-26NINGBO BOHAI SHENHENG TECH CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
NINGBO BOHAI SHENHENG TECH CO LTD
Filing Date
2026-02-06
Publication Date
2026-05-26

AI Technical Summary

Technical Problem

Existing underwater acoustic detection technologies have low detection accuracy when dealing with low-altitude, slow-moving, and stealthy targets, making it difficult to accurately reconstruct the target's outline features.

Method used

By constructing a time-varying channel background model, utilizing dynamic channel impulse response analysis, and combining it with a deep learning network, a refined image of the abnormal region is generated. The target contour is then determined through differential features, reducing environmental noise interference and improving detection accuracy.

Benefits of technology

It significantly improves the detection accuracy of underwater targets, reduces the probability of false alarms, shortens the time from target detection to contour confirmation, and meets the detection requirements of high-speed moving targets.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121661481B_ABST
    Figure CN121661481B_ABST
Patent Text Reader

Abstract

This invention discloses an underwater target detection method based on dynamic channel impulse response analysis, relating to the field of underwater target detection and recognition technology based on dynamic channel impulse response analysis. The method includes the following steps: determining anomaly regions based on the difference between the instantaneous channel impulse response and the reference channel impulse response; generating a refined image of the anomaly regions by adaptively filling the pixels in the anomaly regions and adjusting the sampling step size of the anomaly regions; obtaining difference features by performing differential extraction on the channel impulse response of the refined image and the corresponding reference channel impulse response; and determining the target contour based on the difference features. This application determines the target difference features by matching the channel impulse response of the refined image with the reference channel impulse response. Based on the difference features, a reconstructed contour that is very similar to the true contour of the target can be obtained, thus highly restoring the geometric shape of the underwater target and significantly improving detection accuracy.
Need to check novelty before this filing date? Find Prior Art