基于连续帧与注意力机制的水下小目标自主检测算法

Underwater Small Target Detection Based on Continuous Frame and Attention Mechanism

摘要:

针对声呐对水下小目标探测时目标特征少,常规目标检测算法性能不佳的问题,提出了一种以YOLOv5目标识别算法为基础的连续帧识别改进算法。该算法通过连续帧数据提取模块以及轻量的通道空间注意力模块,提取声呐连续帧信息,提升了YOLOv5算法的识别能力。湖上前视声呐时序数据集算法验证试验表明,在几乎不增加推理时间的前提下,改进算法的平均检测精度比YOLOv5算法提升了13.7%。该改进算法预期可在水下小目标自主检测任务中应用。

Abstract:

In order to solve the problem of less target characteristics and poor performance of conventional target detection algorithm for underwater small target detection by Sonar, a continuous frame recognition modification based on? YOLOv5 is proposed. The algorithm through a lightweight channel and space attention module to extract sonar continuous frame information which improved the ability of YOLOv5. The lake experimental results based on the forward-looking sonar time series dataset show that the accuracy of the algorithm is improved by 13.7% and the reasoning time is basically unchanged. The improved algorithm is expected to be applied to the autonomous detection of underwater small targets.