Passive target detection under low signal-to-noise ratio is essential in various underwater circumstances, but difficult as lacking of targets’ prior information, special propagation characteristics and effect of background noise. Based on the time–frequency representation, we develop a narrow band time–frequency space matched method. The time–frequency matrix is derived based on the ray theory, the warping expression, and the pooling function. Then the sliding cumulative convolution of the matrix obtains a higher signal to noise ratio output. The likelihood ratio test detector is derived for passive target detection. Analysis and the ocean experiment results show that the proposed approach outperforms other passive detectors.

低信噪比下的无源目标检测在各种水下环境中是必不可少的，但由于缺乏目标的先验信息、特殊的传播特性和背景噪声的影响而困难重重。基于时频表示，我们开发了一种窄带时频空间匹配方法。时频矩阵是基于射线理论、翘曲表达式和池化函数推导出来的。然后矩阵的滑动累积卷积得到更高的信噪比输出。似然比测试检测器用于被动目标检测。分析和海洋实验结果表明，所提出的方法优于其他无源探测器。