The Dual-Band SAR Image Fusion-Based Foliage-Penetrating Target Detection Method

Co-registered dual-band (Ku-band and P-band) LSAR image patches with size 1024×1024 were recorded by our developed LSAR system in this article. As LSAR images shown in Fig. 12, the imaging scene is a wetland with a water area and vegetation area.

Breakthrough in Foliage-Penetrating Radar Could Transform Military Surveillance and Humanitarian Operations

New dual-band SAR fusion technique dramatically improves detection of hidden targets while reducing false alarms and computational costs

Chinese researchers have developed a revolutionary approach to detecting targets concealed beneath dense foliage using synthetic aperture radar (SAR) technology, potentially transforming military surveillance, counter-terrorism operations, and humanitarian demining efforts. The breakthrough, published in IEEE Transactions on Geoscience and Remote Sensing, represents a significant advancement in foliage-penetrating (FOPEN) radar capabilities.

The research team from the National University of Defense Technology in Changsha, China, led by Haonan Zhang and Associate Professor Daoxiang An, developed what they call the Dual-Band SAR Image Fusion-Based Foliage-Penetrating Target Detection (DBIFFPTD) method. This innovative technique combines data from both low-frequency and high-frequency radar systems to create synthetic images that can reveal hidden military vehicles, weapons, and other targets while dramatically reducing false alarms.

The Challenge of Seeing Through Foliage

Foliage-penetrating radar has been a holy grail of military technology since the Vietnam War, when the dense tropical forests provided insurgent forces with natural concealment. Low-frequency radar systems, operating in P-band (around 65 cm wavelength) or L-band (around 23 cm wavelength), can penetrate vegetation to detect hidden objects. However, these systems face a critical limitation: they also detect exposed targets, creating numerous false alarms that make it difficult to distinguish between concealed and visible objects.

"The fundamental challenge is that low-band SAR images present both foliage-penetrating targets and exposed strong-scattering targets," explains Dr. An. "This makes direct detection problematic due to high false alarm rates."

Traditional approaches have attempted to solve this by using dual-band systems that combine low-frequency penetrating radar with high-frequency systems that primarily detect surface features. However, the different statistical properties and texture characteristics of these two image types have made direct comparison impossible.

A Revolutionary Fusion Approach

The Chinese team's breakthrough lies in their novel texture-transferring method based on dual-band SAR image fusion (DBIF). Instead of using computationally expensive deep learning approaches like cycle-consistent generative adversarial networks (CGANs), which can take over 8 hours to process and often produce discontinuous results, the DBIFFPTD method employs a more efficient fusion technique.

The process works by decomposing dual-band SAR images using a mathematical framework called nonsubsampled shearlet transform (NSST), which separates images into different frequency components. Low-frequency components are fused using nonnegative matrix factorization (NMF), while high-frequency components are combined using a method based on modified Laplacian energy that preserves edge and texture information.

"Our approach transfers texture information from high-band SAR images to low-band SAR images to generate synthetic high-band images," says Zhang. "This allows us to perform difference operations that highlight only the foliage-penetrating targets while eliminating false alarms from exposed objects."

Impressive Performance Results

The researchers tested their method on both linear SAR (LSAR) and circular SAR (CSAR) systems using real-world data. In experiments detecting seven vehicles hidden under trees, the DBIFFPTD method successfully identified all targets with only a few small false alarms, while traditional methods struggled with numerous false detections.

Perhaps most significantly, the new approach reduces processing time from over 8 hours (required by CGAN-based methods) to just 105 seconds, while still outperforming conventional double-parameter constant false alarm rate (DCFAR) detectors that require only 45 seconds but produce many more false alarms.

The method's effectiveness was demonstrated across different scenarios, from farmland containing hidden vehicles to urban street environments with concealed street lamps. Structural similarity measurements showed that the DBIFFPTD method achieved similarity values of 0.97 compared to just 0.036 for competing approaches.

Broader Applications and Future Impact

The implications of this breakthrough extend far beyond military applications. The Tactical Reconnaissance and Counter-Concealment-Enabled Radar (TRACER) system, recently deployed by the U.S. Army, demonstrates the growing importance of foliage-penetrating radar for counter-terrorism operations. SRI International has been pioneering FOPEN technology for humanitarian applications, including detection of buried landmines and unexploded ordnance.

Current ultra-wideband (UWB) SAR systems operating at frequencies between 100 MHz and 3 GHz offer excellent penetration capabilities but often suffer from poor spatial resolution. The Chinese team's fusion approach helps overcome this limitation by combining the penetration capabilities of low-frequency systems with the high-resolution imagery of higher-frequency radars.

Technical Innovation in Context

The research builds on decades of development in synthetic aperture radar technology. The U.S. Army Research Laboratory's railSAR system, developed in the early 1990s, operated across a 950 MHz-wide band and provided some of the first demonstrations of foliage-penetrating capabilities. However, these early systems required about 80 hours to collect complete high-resolution data.

Recent advances in deep learning have led to new approaches for SAR image interpretation. Researchers have successfully applied convolutional neural networks (CNNs) and generative adversarial networks (GANs) to low-resolution FOPEN SAR images for automatic target recognition. However, these methods often require extensive computational resources and training data.

The DBIFFPTD method represents a middle ground, offering significant improvements over traditional approaches while remaining computationally practical for real-world deployment.

Looking Forward

As radar technology continues to advance, the ability to see through natural concealment becomes increasingly important for both military and civilian applications. Modern SAR systems can achieve resolutions as high as 25 cm and provide all-weather monitoring capabilities, making them invaluable for emergency response, defense intelligence, and agricultural monitoring.

The Chinese research team's innovation demonstrates how mathematical and signal processing advances can dramatically improve radar capabilities without requiring entirely new hardware systems. By efficiently combining information from existing dual-band radar systems, the DBIFFPTD method opens new possibilities for enhanced surveillance and reconnaissance capabilities.

Future research directions include applying the technique to even broader frequency ranges and exploring its effectiveness in different terrain types and weather conditions. The team also plans to investigate applications in unmanned aerial vehicle (UAV) platforms, which could provide more flexible deployment options for both military and humanitarian missions.

As global security challenges continue to evolve, innovations like DBIFFPTD represent crucial advances in the ongoing effort to maintain technological superiority while supporting humanitarian objectives such as landmine detection and disaster response.

---

Sources

1. Zhang, H., An, D., Li, J., Chen, L., Feng, D., Song, Y., & Zhou, Z. (2024). The Dual-Band SAR Image Fusion-Based Foliage-Penetrating Target Detection Method. IEEE Transactions on Geoscience and Remote Sensing, 62, 5226513. https://ieeexplore.ieee.org/document/
2. "Foliage Penetration Radar (2024)" IEEE Aerospace and Electronic Systems Society. https://ieee-aess.org/presentation/lecture/foliage-penetration-radar-2024
3. Uppal, R. (2023, November 14). "Beyond the Green Canopy: Synthetic Aperture Radar (SAR) and its Foliage Penetration Capabilities." International Defense Security & Technology. https://idstch.com/security/beyond-the-green-canopysynthetic-aperture-radar-sar-and-its-foliage-penetration-capabilities/
4. NASA Earthdata. (2025, April 16). "Synthetic Aperture Radar (SAR)." https://www.earthdata.nasa.gov/learn/earth-observation-data-basics/sar
5. Kechagias-Stamatis, O., & Aouf, N. (2021). "Automatic Target Recognition for Low Resolution Foliage Penetrating SAR Images Using CNNs and GANs." Remote Sensing, 13(4), 596. https://www.mdpi.com/2072-4292/13/4/596
6. "Adaptive target detection in foliage-penetrating SAR images using alpha-stable models." IEEE Xplore. https://ieeexplore.ieee.org/document/806628/
7. "Foliage-penetrating reconnaissance radar to support counter-terrorism operations." (2025, June 30). Military Embedded Systems. https://militaryembedded.com/radar-ew/sensors/foliage-penetrating-reconnaissance-radar-to-support-counter-terrorism-operations
8. Alves, W.A.L., et al. (2010). "Sense-Through-Foliage target detection using UWB radar sensor networks." Pattern Recognition Letters, 31(9), 107-114. https://www.sciencedirect.com/science/article/abs/pii/S0167865510000863
9. "75 Years of Innovation: Foliage-Penetrating Radar technologies (FOLPEN)." (2025, February 5). SRI International. https://www.sri.com/75-years-of-innovation/75-years-of-innovation-foliage-penetrating-radar-technologies-folpen/
10. "RailSAR." (2024, December 14). Wikipedia. https://en.wikipedia.org/wiki/RailSAR
11. "What Is SAR Imagery? Introduction To Synthetic Aperture Radar." (2024, May 23). European Space Imaging. https://www.euspaceimaging.com/blog/2024/04/05/what-is-sar-imagery/
12. "SAR image target recognition based on NMF feature extraction and Bayesian decision fusion." IEEE Xplore. https://ieeexplore.ieee.org/document/5602633/
13. Bai, X., et al. (2015). "Multi-sensor image enhanced fusion algorithm based on NSST and top-hat transformation." Optics Communications, 347, 1-17. https://www.sciencedirect.com/science/article/abs/pii/S0030402615008797
14. Huang, J., et al. (2023). "An NSST-Based Fusion Method for Airborne Dual-Frequency, High-Spatial-Resolution SAR Images." IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 16, 4362-4370. https://ieeexplore.ieee.org/document/10109765/
15. The Dual-Band SAR Image Fusion-Based Foliage-Penetrating Target Detection Method | IEEE Journals & Magazine | IEEE Xplore