Radar Image Super-Resolution Without Additional Radar Hardware
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High-frequency radar systems require numerous transmitters and receivers to achieve high-resolution imaging, increasing cost and complexity, while existing methods for converting low-resolution radar images to high-resolution images are inefficient.
Innovation Solution
A device and method utilizing RF simulation and deep learning to train a super-resolution model, converting low-resolution radar images into high-resolution images using generative adversarial networks, transformers, and convolutional neural networks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-frequency radar with short wavelength is used to obtain high-resolution image, then image resolution is improved, but the number of transmitters and receivers increases, increasing system cost and complexity
Solution Approach 1:
The patent creates virtual high-resolution radar images by training a deep learning model on simulated radar data. Instead of physically deploying more transmitters and receivers, the system copies the effect of high-resolution imaging through computational generation, where the trained model synthesizes high-resolution images from low-resolution inputs, eliminating the need for additional hardware components
Solution Approach 2:
The patent replaces the mechanical approach of increasing hardware (more transmitters and receivers) with a computational approach using deep learning models. The mechanical system of physical radar components is substituted with an information-processing system that uses trained neural networks to achieve high-resolution imaging through software rather than hardware expansion
2Measurement precision
If high-frequency radar with short wavelength is used to obtain high-resolution image, then image resolution is improved, but system cost increases due to more transmitters and receivers
Solution Approach 1:
The patent creates virtual high-resolution radar images by training a deep learning model on simulated radar data. Instead of physically deploying more transmitters and receivers, the system copies the effect of high-resolution imaging through computational generation, where the trained model synthesizes high-resolution images from low-resolution inputs, eliminating the need for additional hardware components
Solution Approach 2:
The patent uses computational resources and trained models as a cheaper alternative to expensive physical hardware. Rather than investing in costly additional transmitters and receivers, the system uses software-based super-resolution techniques that can be deployed and updated without physical hardware changes, reducing overall system cost
Data Source
AI summary
Provided are a device and method for radar image super-resolution. The device includes a memory configured to store at least one instruction and a processor configured to execute the at least one instruction stored in the memory. The processor generates a low-resolution radar image of a target, generates a high-resolution radar image of the target, trains a super-resolution model for performing super-resolution on radar images on the basis of the low-resolution radar image and the high-resolution radar image, and performs super-resolution on a target radar image using the trained super-resolution model.