陈灿灿, 夏润秋, 刘洋, 刘力双, 陈青山. 层次卷积滤波红外弱小目标检测方法[J]. 应用光学, 2023, 44(4): 826-833. DOI: 10.5768/JAO202344.0403003
引用本文: 陈灿灿, 夏润秋, 刘洋, 刘力双, 陈青山. 层次卷积滤波红外弱小目标检测方法[J]. 应用光学, 2023, 44(4): 826-833. DOI: 10.5768/JAO202344.0403003
CHEN Cancan, XIA Runqiu, LIU Yang, LIU Lishuang, CHEN Qingshan. Hierarchical convolution filtering method for infrared dim small target detection[J]. Journal of Applied Optics, 2023, 44(4): 826-833. DOI: 10.5768/JAO202344.0403003
Citation: CHEN Cancan, XIA Runqiu, LIU Yang, LIU Lishuang, CHEN Qingshan. Hierarchical convolution filtering method for infrared dim small target detection[J]. Journal of Applied Optics, 2023, 44(4): 826-833. DOI: 10.5768/JAO202344.0403003

层次卷积滤波红外弱小目标检测方法

Hierarchical convolution filtering method for infrared dim small target detection

  • 摘要: 针对单帧复杂背景红外图像点目标检测算法存在复杂背景下处理效果不理想、处理时间长的问题,提出了一种层次卷积滤波检测算法。主要分为两个部分:第一,根据红外小目标特性,设计一种层次卷积滤波的算子,对图像进行滤波处理,实现图像中小目标的增效和背景抑制的效果;第二,采用基于最大值的自适应阈值方法,对图像进行二值化操作,过滤背景杂波,最终提取到待检测的目标。在大量不同背景红外图像中进行实验,论文算法在背景抑制因子和信噪比增益的性能量化结果上优于现有5种典型红外弱小目标检测算法的性能结果,且平均处理时间仅为高斯拉普拉斯(Laplacian of Gaussian,LoG)滤波算法的30.42%。通过实验对比,表明该层次卷积滤波算法可以有效解决在不同复杂背景下的红外图像中对小目标检测的问题。

     

    Abstract: Aiming at the problems of unsatisfactory processing effect and long processing time in point target detection algorithm of single-frame infrared image with complex background, a hierarchical convolution filtering detection algorithm was proposed. It was mainly divided into two parts: firstly, according to the characteristics of small infrared targets, a hierarchical convolution filtering operator was designed to filter the image, so as to achieve the effect of efficiency increase and background suppression of small targets in the image. Secondly, the adaptive threshold method based on the maximum value was used to binarize the image to filter the background clutter, and finally extracted the target to be detected. Experiments in a large number of infrared images with different backgrounds show that the performance quantization results of background suppression factor and signal-to-noise ratio gain of the algorithm are better than that of the existing five typical infrared dim and small target detection algorithms, and the average processing time is only 30.42% of Laplacian of Gaussian (LoG) filtering algorithm. Through the experimental comparison, the hierarchical convolution filtering method can effectively solve the problem of small target detection in infrared images under different complex backgrounds.

     

/

返回文章
返回