基于多特征的红外双波段目标检测融合算法

Infrared Dual-band Target Detecting Fusion Algorithm Based on Multiple Features

红外目标检测在军事领域和民用领域发挥着重要的作用,得到了广泛的研究,但对于利用双波段图像对目标进行检测的研究较少。为了充分发挥双波段图像在目标检测性能上的优势,对红外双波段图像的检测结果进行深入分析,提出了一种基于多特征的红外双波段目标检测融合算法。本文提出的融合算法,利用基于深度学习的多特征融合网络对双波段图像的检测结果进行处理,充分挖掘目标的特征信息,自适应地选择单波段的检测结果作为输出,得到最终的决策级融合检测结果。实验结果表明:与使用单波段图像进行目标检测相比,本文提出的基于多特征的红外双波段融合算法,可以有效利用不同波段的信息,提高检测性能,充分发挥红外目标探测设备的优势。

Infrared target detection algorithms play important roles in the military and civilian fields and have been widely studied.However,relatively few studies have been conducted on the use of dual-band images for targeted detection.To fully utilize the advantages of dual-band images in target detection,a fusion algorithm based on multiple features of infrared dual-band images was proposed through an in-depth analysis of the detection results.The proposed fusion algorithm utilizes a deep learning-based multi-feature fusion network to process the detection results of dual-band images,fully mine the feature information of the target,adaptively select the detection results of a single band as the output,and obtain the final decision-level fusion detection results.The experimental results show that,compared with using single-band images for object detection,the proposed infrared dual-band fusion algorithm based on multiple features can effectively utilize information from different bands,improve the detection performance,and fully leverage the advantages of infrared object detection equipment.