GPR-TSBiNet: An Information Gradient Enrichment Model for GPR B-Scan Small Target Detection

Accurate detection of underground grounding lines remains a significant technical challenge due to their deep burial and small cross-sectional dimensions, which cause signal scattering in heterogeneous soil media. This results in blurred features in GPR B-scan images, impeding reliable target identi...

Full description

Saved in:
Bibliographic Details
Main Authors: Chongqin Wang, Yi Guan, Minghe Chi, Feng Shen, Zhilong Yu, Qingguo Chen, Chao Chen
Format: Article
Language:English
Published: MDPI AG 2025-04-01
Series:Sensors
Subjects:
Online Access:https://www.mdpi.com/1424-8220/25/7/2223
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849730084643012608
author Chongqin Wang
Yi Guan
Minghe Chi
Feng Shen
Zhilong Yu
Qingguo Chen
Chao Chen
author_facet Chongqin Wang
Yi Guan
Minghe Chi
Feng Shen
Zhilong Yu
Qingguo Chen
Chao Chen
author_sort Chongqin Wang
collection DOAJ
description Accurate detection of underground grounding lines remains a significant technical challenge due to their deep burial and small cross-sectional dimensions, which cause signal scattering in heterogeneous soil media. This results in blurred features in GPR B-scan images, impeding reliable target identification. To address this limitation, we propose GPR-TSBiNet, an architecture incorporating two key model innovations. We introduce GPR-Transformer (GPR-Trans), a multi-branch backbone network specifically designed for GPR B-scan processing. In the neck stage, we develop the Spatial-Depth Converted Bidirectional Feature Pyramid Network (SC-BiFPN), which integrates SPD-ADown to mitigate feature loss caused by traditional pooling-based downsampling. We employ Shape-IoU as the loss function to enhance boundary detail preservation for small targets. Comparative experiments demonstrate that GPR-TSBiNet outperforms state-of-the-art (SOTA) models YOLOv11 and YOLOv10 in detection accuracy, achieving an AP0.5 improvement of 11.6% over YOLOv11X and 27.4% over YOLOv10X. Notably, the model improves small-target APsmall to 49.4 ± 0.7%, representing a 13.4% increase over the SOTA YOLOv11 model. Finally, real-world GPR validation experiments are conducted, confirming that GPR-TSBiNet provides a reliable solution for underground grounding line detection in GPR-based target recognition.
format Article
id doaj-art-0c4eab5bc8dc488eb8424e5fa47424d5
institution DOAJ
issn 1424-8220
language English
publishDate 2025-04-01
publisher MDPI AG
record_format Article
series Sensors
spelling doaj-art-0c4eab5bc8dc488eb8424e5fa47424d52025-08-20T03:08:59ZengMDPI AGSensors1424-82202025-04-01257222310.3390/s25072223GPR-TSBiNet: An Information Gradient Enrichment Model for GPR B-Scan Small Target DetectionChongqin Wang0Yi Guan1Minghe Chi2Feng Shen3Zhilong Yu4Qingguo Chen5Chao Chen6School of Electrical and Electronic Engineering, Harbin University of Science and Technology, Harbin 150080, ChinaSchool of Electrical and Electronic Engineering, Harbin University of Science and Technology, Harbin 150080, ChinaSchool of Electrical and Electronic Engineering, Harbin University of Science and Technology, Harbin 150080, ChinaSchool of Instrumentation Science and Engineering, Harbin Institute of Technology, Harbin 150001, ChinaSchool of Electrical and Electronic Engineering, Harbin University of Science and Technology, Harbin 150080, ChinaSchool of Electrical and Electronic Engineering, Harbin University of Science and Technology, Harbin 150080, ChinaCollege of Materials Science and Engineering, Taiyuan University of Technology, Taiyuan 030024, ChinaAccurate detection of underground grounding lines remains a significant technical challenge due to their deep burial and small cross-sectional dimensions, which cause signal scattering in heterogeneous soil media. This results in blurred features in GPR B-scan images, impeding reliable target identification. To address this limitation, we propose GPR-TSBiNet, an architecture incorporating two key model innovations. We introduce GPR-Transformer (GPR-Trans), a multi-branch backbone network specifically designed for GPR B-scan processing. In the neck stage, we develop the Spatial-Depth Converted Bidirectional Feature Pyramid Network (SC-BiFPN), which integrates SPD-ADown to mitigate feature loss caused by traditional pooling-based downsampling. We employ Shape-IoU as the loss function to enhance boundary detail preservation for small targets. Comparative experiments demonstrate that GPR-TSBiNet outperforms state-of-the-art (SOTA) models YOLOv11 and YOLOv10 in detection accuracy, achieving an AP0.5 improvement of 11.6% over YOLOv11X and 27.4% over YOLOv10X. Notably, the model improves small-target APsmall to 49.4 ± 0.7%, representing a 13.4% increase over the SOTA YOLOv11 model. Finally, real-world GPR validation experiments are conducted, confirming that GPR-TSBiNet provides a reliable solution for underground grounding line detection in GPR-based target recognition.https://www.mdpi.com/1424-8220/25/7/2223feature extractiontransformersGPRdeep learningoptical imaging
spellingShingle Chongqin Wang
Yi Guan
Minghe Chi
Feng Shen
Zhilong Yu
Qingguo Chen
Chao Chen
GPR-TSBiNet: An Information Gradient Enrichment Model for GPR B-Scan Small Target Detection
Sensors
feature extraction
transformers
GPR
deep learning
optical imaging
title GPR-TSBiNet: An Information Gradient Enrichment Model for GPR B-Scan Small Target Detection
title_full GPR-TSBiNet: An Information Gradient Enrichment Model for GPR B-Scan Small Target Detection
title_fullStr GPR-TSBiNet: An Information Gradient Enrichment Model for GPR B-Scan Small Target Detection
title_full_unstemmed GPR-TSBiNet: An Information Gradient Enrichment Model for GPR B-Scan Small Target Detection
title_short GPR-TSBiNet: An Information Gradient Enrichment Model for GPR B-Scan Small Target Detection
title_sort gpr tsbinet an information gradient enrichment model for gpr b scan small target detection
topic feature extraction
transformers
GPR
deep learning
optical imaging
url https://www.mdpi.com/1424-8220/25/7/2223
work_keys_str_mv AT chongqinwang gprtsbinetaninformationgradientenrichmentmodelforgprbscansmalltargetdetection
AT yiguan gprtsbinetaninformationgradientenrichmentmodelforgprbscansmalltargetdetection
AT minghechi gprtsbinetaninformationgradientenrichmentmodelforgprbscansmalltargetdetection
AT fengshen gprtsbinetaninformationgradientenrichmentmodelforgprbscansmalltargetdetection
AT zhilongyu gprtsbinetaninformationgradientenrichmentmodelforgprbscansmalltargetdetection
AT qingguochen gprtsbinetaninformationgradientenrichmentmodelforgprbscansmalltargetdetection
AT chaochen gprtsbinetaninformationgradientenrichmentmodelforgprbscansmalltargetdetection