J-MSF: 一种新的多通道多尺度红外弱小目标检测算法

针对经典的基于深度学习的红外弱小目标检测算法存在目标信息在高层感受野消失导致无法检出的问题，提出一种新的基于多通道多尺度特征融合的红外弱小目标检测算法（J-MSF）。首先，该算法提出了一种新的多通道JAnet结构，基于此结构搭建了主干特征提取网络；其次，设计了下降门限式特征金字塔池化结构(DSPP)，并提出了多尺度融合检测策略；最后，设计了高斯损失优化函数。实验结果表明，所提出的算法在“地/空背景下红外图像弱小飞机目标检测跟踪数据集”上的检测效果与YOLOv3、YOLOv4算法对比，检出率、整体AP值分别提升9.07%、9.89%和1.67%、3.16%，提出算法优于目前主流的检测算法，体现出了良好的鲁棒性和适应性，可以有效的应用于红外弱小目标的检测。

J-MSF:A new infrared dim and small target detection algorithm based on multi-channel and multiscale

An novel infrared dim and small target detection algorithm, called J-MSF, based on multi-channel and multi-scale feature fusion was proposed, which solved the problem that the classical infrared dim and small target detection algorithm based on deep learning cannot detect because the target information disappeared in the upper receptive field. Firstly, a new multi-channel Janet structure was proposed to design the J-MSF backbone extraction framework. Secondly, a descending threshold feature pyramid pooling structure (DSPP) was exploited, and a multi-scale fusion detection strategy was conducted. Finally, the Gauss loss optimization function was designed. The experimental results show that the recall rate and the AP value of the proposed algorithm are improved by 9.07%, 9.89% and 1.67%, 3.16%, respectively, compared with those of YOLOv3 and YOLOv4 algorithms in "a dataset for infrared detection and tracking of dim and small aircraft targets underground/air background". The proposed algorithm can be effectively applied to infrared dim and small target detection, shows good robustness and adaptability, and is better than the state of the art algorithms.