High-Resolution Through-Wall Imaging Using Data Fusion and Reasoning

Through-wall radar has been very pertinent to a variety of civilian and military services because of its ability to detect and sense through the wall obstacles. However, to maintain the penetrating ability, most of the existing TWR systems work at L/S band with limited bandwidth and thus can only ge...

Full description

Saved in:
Bibliographic Details
Main Authors: Chen, Zihan, Zeng, Xiaolu, Yang, Xiaopeng, Zhao, Jiarong, Gong, Junbo
Format: Conference Proceeding
Language:English
Subjects:
Cognition
data fusion and reasoning
Data integration
generative adversarial network
Image resolution
outline reconstruction
Radar
Radar detection
Radar imaging
Shape
Through-wall radar imaging
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Through-wall radar has been very pertinent to a variety of civilian and military services because of its ability to detect and sense through the wall obstacles. However, to maintain the penetrating ability, most of the existing TWR systems work at L/S band with limited bandwidth and thus can only generate a very crude blob of the target, whose resolution is not easy-to-use for many practical applications. To address this issue, this paper proposes a novel high resolution TWR imaging system by deep learning-based data fusion and reasoning techniques. First, we devise an image-reasoning module by fusing TWR and optical images with generative adversarial networks. Then, in the online phase, the low-resolution TWR image is fed into the image-reasoning module for resolution improvement. Extensive simulations and experiments demonstrate that the proposed method can successfully reconstruct the outline of an object rather than just a blob, which greatly eases the end user to interpret and thus facilitating more applications.
ISSN:2379-190X
DOI:10.1109/ICASSP48485.2024.10448456