Diffusion 기반 클러터 제거 고해상도 거리-도플러 맵 생성

홍창인¹^{, *} ; 김홍락² ; 박성호² ; 김민각¹ ; 조성국¹

1LIG넥스원 PGM 탐색기연구소 연구원
2LIG넥스원 PGM 탐색기연구소 수석연구원

Diffusion-based High Resolution Range-Doppler Map Generation with Clutter Removal

Chang-In Hong¹^{, *} ; Hong-Rak Kim² ; Sung-Ho Park² ; Min-Gak Kim¹ ; Seong-Guk Cho¹

1Research engineer, PGM RF&IIR Seeker R&D Lab, LIG Nex1
2Chief research engineer, PGM RF&IIR Seeker R&D Lab, LIG Nex1

Correspondence to: ^*Chang-In Hong PGM RF & IIR SEEKER R&D Lab, LIG Nex1 207, Mabuk-ro, Giheung-gu, Yongin-si, Gyeonggi-do, Republic of Korea Tel: +82-31-525-0305 E-mail: changin.hong@lignex1.com

초록

본 논문에서는 비행 환경의 유도무기에 탑재된 탐색기가 표적을 탐지 및 추적하는 과정에서 필요한 정보인 range-Doppler map을 생성형 모델인 DDPM(Denoising Diffusion Probabilistic Models) 기반으로 생성하는 방법을 제시한다. 비행 시험에 의해 한정적으로 획득 가능한 실제 RD map과 유사한 가상 RD map을 대량 생성하여 탐색기의 동작 로직을 검증하기 위한 데이터를 확보한다. 추가적으로 저해상도 RD map 픽셀 사이의 이미지를 생성형 모델을 통해 예측하여 고해상도 RD map을 생성한다.

Abstract

This paper proposed a method for generating a range-Doppler map based on the DDPM(Denoising Diffusion Probabilistic Models), a generative model, which is the necessary information for a seeker mounted on a guided weapon in a flight environment to detect and track a target. By generating a large amount of virtual RD maps similar to the actual RD map that can be obtained only by flight tests, data for verifying the operational logic of the seeker is obtained. Additionally, images between low-resolution RD map pixels are predicted through generative models to generate high-resolution RD maps.

Keywords:

FMCW Radar, Denoising Diffusion Probabilistic Models, Captive Flight Test, Range-Doppler Map, High Resolution

키워드:

FMCW 레이다, DDPMs, 항공기 탑재 시험, 거리-도플러 맵, 고해상도

Acknowledgments

이 논문은 2024년도 한국해군과학기술학회 동계학술대회 발표 논문임

References

Son, Chang-Hee, “Estimation of Ground Clutter Reflectivity Based on the CFT(Captive Flight Test),” Journal of the Korea Institute of Military Science and Technology, Vol. 9, No. 2, pp. 87-95, 2006.
T. Jeong and S. Lee, “Resource-Efficient Range-Doppler Map Generation Using Deep Learning Network for Automotive Radar Systems,” in IEEE Access, vol. 11, pp. 55965-55977, 2023. [https://doi.org/10.1109/ACCESS.2023.3282688]
Min Jong Lee, Dong Woo Kim, “Robust Evaluation of Diffusion-Based Adversarial Purification”, Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2023, pp. 134-144. [https://doi.org/10.1109/ICCV51070.2023.00019]
Ho, J., Chen, X., Srinivas, A., Duan, Y., & Abbeel, P., “Denoising Diffusion Probabilistic Models.” arXiv preprint arXiv, 11239, 2020.
I. Roldan, C. del-Blanco, Á. Quevedo, F. Urzaiz, J. Menoyo, A. López et al., “Dopplernet: A Convolutional Neural Network for Recognising Targets in Real Scenarios Using a Persistent Range–Doppler Radar,” IET Radar, Sonar & Navigation, Vol. 14, No. 4, pp. 593-600, 2020. [https://doi.org/10.1049/iet-rsn.2019.0307]