구조적 희소성 기반 압축 센싱 알고리즘을 통한 측면주사소나 영상의 비균일 잡음 제거 = Non-homogeneous noise removal for side scan sonar images using a structural sparsity based compressive sensing algorithm

다국어 초록 (Multilingual Abstract)

The quality of side scan sonar images is determined by the frequency of a sonar. A side scan sonar with a low frequency creates low-quality images. One of the factors that lead to low quality is a high-level noise. The noise is occurred by the underwater environment such as equipment noise, signal interference and so on. In addition, in order to compensate for the transmission loss of sonar signals, the received signal is recovered by TVG (Time-Varied Gain), and consequently the side scan sonar images contain non-homogeneous noise which is opposite to optic images whose noise is assumed as homogeneous noise. In this paper, the SSCS (Structural Sparsity based Compressive Sensing) is proposed for removing non-homogeneous noise. The algorithm incorporates both local and non-local models in a structural feature domain so that it guarantees the sparsity and enhances the property of non-local self-similarity. Moreover, the non-local model is corrected in consideration of non-homogeneity of noises. Various experimental results show that the proposed algorithm is superior to existing method.

국문 초록 (Abstract)

측면주사소나 영상의 화질은 소나 운용 주파수의 영향을 받는다. 저주파 측면주사소나 장비로 얻는 영상은 저화질 영상이며, 잡음이 화질 저하의 요소 중 하나가 된다. 균일한 잡음을 가정하는 광학 영상과는 달리. 측면주사소나 데이터의 잡음은 해양 환경(장비 소음, 신호 간섭 등)에 의해 발생한다. 또한 소나 신호의 전달 손실을 보상하고자 시간변환이득(Time-Varied Gain, TVG)을 수행하며, 이로 인해 측면주사소나 영상에 비균일 잡음이 생성된다. 본 논문에서는 측면주사소나 영상에 포함된 비균일 잡음을 제거하는 구조적 희소성에 기반한 압축 센싱 알고리즘 (Structural Sparsity based Compressive Sensing, SSCS)을 제안한다. 영상의 구조적 특징 도메인에서 국부적 및 비국부적 모델링을 동시에 구현하여 계수의 희소성을 보장하면서 비국부적 자가 유사성을 강화한다. 그리고 잡음의 비균일성을 고려하여 비국부적 모델링을 보상한다. 다양한 모의 실험을 통해 제안한 알고리즘의 우수성을 입증한다.

참고문헌 (Reference)

1 M. Aharon, "The {K-SVD}: An algorithm for designing of overcomplete dictionaries for sparse representations" 54 : 4311-4322, 2006

2 W. Dong, "Sparsity-based image denoising via dictionary learning and structural clustering" 457-464, 2011

3 T. Boas, "Shrinkage function and Its applications in matrix approximations" 32 : 163-171, 2017

4 G. Zhao, "Rotation invariant image and video description with local binary pattern features" 21 : 1-13, 2010

5 H. Abdi, "Principal component analysis" 2 : 433-459, 2010

6 O. Tuzel, "Pedestrian detection via classification on Riemannian manifolds" 30 : 1713-1727, 2008

7 J. Mairal, "Non-local sparse models for image restoration" 2272-2279, 2009

8 A. Buades, "Neighborhood filters and pde's" CMLA 2005

9 M. Moszynski, "Increasing the accuracy of time-varied-gain in digital echosounders" 88 : 814-817, 2002

10 J. Lv, "Image laplace denoising based on sparse representation" 373-377, 2015

11 J. Portilla, "Image denoising using scale mixtures of Gaussians in thewavelet domain" 12 : 1338-1351, 2003

12 K. Dabov, "Image denoising by sparse 3-d transform-domain collaborative filtering" 16 : 2080-2095, 2007

13 M. Yang, "Gabor feature based sparse representation for face recognition with Gabor occlusion dictionary" 448-461, 2010

14 L. W. Kang, "Feature-based sparse representation for image similarity assessment" 13 : 1019-1030, 2011

15 P. Chatterjee, "Clustering-based denoising with locally learned dictionaries" 18 : 1438-1451, 2009

16 Y. Chen, "Analyze the sonar image according to the frequency and altitude of side scan sonar" KSNVE 308-, 2017

17 T. Kanungo, "An efficient k-means clustering algorithm: analysis and implementation" 24 : 881-892, 2002

18 L. Cayton, "Algorithms for manifold learning" Univ. of California at San Diego 973-980, 2005

19 A. Buades, "A non-local algorithm for image denoising" 2 : 60-65, 2005