Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

SE-BSFV: Online Subspace Learning based Shadow Enhancement and Background Suppression for ViSAR under Complex Background

Shangqu Yan, Chenyang Luo, Yaowen Fu, Wenpeng Zhang, Wei Yang, Ruofeng Yu

TL;DR

SE-BSFV tackles shadow enhancement and background suppression in ViSAR by modeling the data as a low-rank background plus sparse moving-shadow foreground while characterizing residuals with a Gaussian Mixture Distribution. The method employs online subspace learning to update subspace and mixture parameters frame-by-frame via EM, followed by SURF-based registration and ADMM refinement to remove strong scatterers. Empirical results on the SNL ViSAR dataset show SE-BSFV markedly improves shadow saliency and detection performance, with favorable processing efficiency compared to existing pre-processing approaches and effective gains across traditional and deep detectors. The approach provides a practical, robust pre-processing pipeline that enhances MTD in complex backgrounds and supports real-time-like operation.

Abstract

Video synthetic aperture radar (ViSAR) has attracted substantial attention in the moving target detection (MTD) field due to its ability to continuously monitor changes in the target area. In ViSAR, the moving targets' shadows will not offset and defocus, which is widely used as a feature for MTD. However, the shadows are difficult to distinguish from the low scattering region in the background, which will cause more missing and false alarms. Therefore, it is worth investigating how to enhance the distinction between the shadows and background. In this study, we proposed the Shadow Enhancement and Background Suppression for ViSAR (SE-BSFV) algorithm. The SE-BSFV algorithm is based on the low-rank representation (LRR) theory and adopts online subspace learning technique to enhance shadows and suppress background for ViSAR images. Firstly, we use a registration algorithm to register the ViSAR images and utilize Gaussian mixture distribution (GMD) to model the ViSAR data. Secondly, the knowledge learned from the previous frames is leveraged to estimate the GMD parameters of the current frame, and the Expectation-maximization (EM) algorithm is used to estimate the subspace parameters. Then, the foreground matrix of the current frame can be obtained. Finally, the alternating direction method of multipliers (ADMM) is used to eliminate strong scattering objects in the foreground matrix to obtain the final results. The experimental results indicate that the SE-BSFV algorithm significantly enhances the shadows' saliency and greatly improves the detection performance while ensuring efficiency compared with several other advanced pre-processing algorithms.

SE-BSFV: Online Subspace Learning based Shadow Enhancement and Background Suppression for ViSAR under Complex Background

TL;DR

SE-BSFV tackles shadow enhancement and background suppression in ViSAR by modeling the data as a low-rank background plus sparse moving-shadow foreground while characterizing residuals with a Gaussian Mixture Distribution. The method employs online subspace learning to update subspace and mixture parameters frame-by-frame via EM, followed by SURF-based registration and ADMM refinement to remove strong scatterers. Empirical results on the SNL ViSAR dataset show SE-BSFV markedly improves shadow saliency and detection performance, with favorable processing efficiency compared to existing pre-processing approaches and effective gains across traditional and deep detectors. The approach provides a practical, robust pre-processing pipeline that enhances MTD in complex backgrounds and supports real-time-like operation.

Abstract

Video synthetic aperture radar (ViSAR) has attracted substantial attention in the moving target detection (MTD) field due to its ability to continuously monitor changes in the target area. In ViSAR, the moving targets' shadows will not offset and defocus, which is widely used as a feature for MTD. However, the shadows are difficult to distinguish from the low scattering region in the background, which will cause more missing and false alarms. Therefore, it is worth investigating how to enhance the distinction between the shadows and background. In this study, we proposed the Shadow Enhancement and Background Suppression for ViSAR (SE-BSFV) algorithm. The SE-BSFV algorithm is based on the low-rank representation (LRR) theory and adopts online subspace learning technique to enhance shadows and suppress background for ViSAR images. Firstly, we use a registration algorithm to register the ViSAR images and utilize Gaussian mixture distribution (GMD) to model the ViSAR data. Secondly, the knowledge learned from the previous frames is leveraged to estimate the GMD parameters of the current frame, and the Expectation-maximization (EM) algorithm is used to estimate the subspace parameters. Then, the foreground matrix of the current frame can be obtained. Finally, the alternating direction method of multipliers (ADMM) is used to eliminate strong scattering objects in the foreground matrix to obtain the final results. The experimental results indicate that the SE-BSFV algorithm significantly enhances the shadows' saliency and greatly improves the detection performance while ensuring efficiency compared with several other advanced pre-processing algorithms.
Paper Structure (20 sections, 36 equations, 9 figures, 5 tables, 1 algorithm)

This paper contains 20 sections, 36 equations, 9 figures, 5 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Previous Work
  3. The SBN-3D-SD Algorithm
  4. The GMD-LRR Algorithm
  5. Methodology
  6. Registration Based on the SURF Algorithm
  7. Online Subspace Learning
  8. The ADMM Processing
  9. Experiments
  10. Data and Experimental Equipment
  11. Implementation Details and Evaluation Metrics
  12. Shadow Enhancement Experiment
  13. Shadow Detection Experiments
  14. Experimental Results and Analysis
  15. Experiment on Shadow Enhancement and Background Suppression
  16. ...and 5 more sections

Figures (9)

  • Figure 1: The flowchart of the SE-BSFV algorithm.
  • Figure 2: Visualization results processed by different pre-processing algorithms. (a) Original images. (b) The MF algorithm. (c) The SAR-BM3D algorithm. (d) The V-BM3D algorithm. (e) The HESE algorithm. (f) The LRSD algorithm. (g) The SBN-3D-SD algorithm. (h) The SE-BSFV algorithm.
  • Figure 3: Enlarged details of red and green boxes from different pre-processing algorithms in Fig.2. (a) Original images. (b) The MF algorithm. (c) The SAR-BM3D algorithm. (d) The V-BM3D algorithm. (e) The HESE algorithm. (f) The LRSD algorithm. (g) The SBN-3D-SD algorithm. (h) The SE-BSFV algorithm.
  • Figure 4: The processing time of the different pre-processing algorithms on the SNL's ViSAR data.
  • Figure 5: The visualization results. (a) Original image detection results. (b) Detection results after the MF processing. (c) Detection results after the SAR-BM3D processing. (d) Detection results after the V-BM3D processing. (e) Detection results after the HESE processing. (f) Detection results after the LRSD processing. (g) Detection results after the SBN-3D-SD processing. (h) Detection results after the SE-BSFV processing.
  • ...and 4 more figures