Journal of the Korea Institute of Information and Communication Engineering (한국정보통신학회논문지)

The Korea Institute of Information and Commucation Engineering (한국정보통신학회)
권호 표지
DOI QR코드

DOI QR Code

Switching Filter based on Noise Estimation in Random Value Impulse Noise Environments

랜덤 임펄스 잡음 환경에서 잡음추정에 기반한 스위칭 필터

  • Bong-Won, Cheon (Dept. of Intelligent Robot Engineering, Pukyong National University) ;
  • Nam-Ho, Kim (School of Electrical Engineering, Pukyong National University)
  • Received : 2022.11.04
  • Accepted : 2022.11.25
  • Published : 2023.01.31
Download PDF
⟨ Previous Next ⟩

Abstract

With the development of IoT technologies and artificial intelligent, diverse digital image equipments are being used in industrial sites. Because image data can be easily damaged by noise while it's obtained with a camera or a sensor and the damaged image has a bad effect on the process of image processing, noise removal is being demanded as preprocessing. In this thesis, for the restoration of image damaged by the noise of random impulse, a switching filter algorithm based on noise estimation was suggested. With the proposed algorithm, noise estimation and error distraction were carried out according to the similarity of the pixel values in the local mask of the image, and a filter was chosen and switched depending on the ratio of noise existing in the local mask. Simulations were conducted to analyze the noise removal performance of the proposed algorithm, and as a result of magnified image and PSNR comparison, it showed superior performance compared to the existing method.

IoT 기술과 인공지능의 발전에 따라 다양한 디지털 영상장비가 산업현장에서 사용되고 있다. 영상 데이터는 카메라 또는 센서에서 취득되는 과정 중 잡음에 훼손되기 쉬우며, 훼손된 영상은 영상처리 과정에서 악영향을 미치기 때문에 전처리 과정으로 잡음제거가 요구되고 있다. 본 논문에서는 랜덤 임펄스 잡음에 훼손된 영상을 복원하기 위해 잡음추정에 기반한 스위칭 필터 알고리즘을 제안하였다. 제안한 알고리즘은 영상의 국부마스크 내부의 화소값의 유사성에 따라 잡음추정과 에러 검출을 진행하였으며, 국부마스크에 존재하는 잡음 비율에 따라 필터를 선택하여 스위칭하였다. 제안하는 알고리즘의 잡음제거 성능을 분석하기 위해 시뮬레이션을 진행하였으며, 확대영상 및 PSNR 비교 결과 기존 방법에 비해 우수한 성능을 나타내었다.

Keywords

References

  1. P. S. V. S. Sridhar and R. Caytiles, "Efficient Cloud Data Hosting Availability," Asia-pacific Journal of Convergent Research Interchange, HSST, ISSN : 2508-9080, vol. 3, no. 2, pp. 11-19, Jun. 2017. DOI: 10.21742/APJCRI.2017.06.02.
  2. B. W. Cheon and N. H. Kim, "AWGN Removal Using Modified Steering Kernel and Image Matching," Applied Sciences, vol. 12, no. 22, pp. 11588, Nov. 2022. DOI: 10.3390/app122211588.
  3. V. Patanavijit and K. Thakulsukanant, "The Novel Denoising Technique Based on 8D Optimal Directions Using Statistical SD for Random-Valued Impulse Noise," in 2018 IEEE 7th Global Conference on Consumer Electronics (GCCE), Nara, Japan, pp. 34-37, 2018. DOI: 10.1109/GCCE. 2018.8574750.
  4. K. Somasundaram and P. Shanmugavadivu, "Impulsive Noise Detection by Second-order Differential Image and Noise Removal using Adaptive Nearest Neighbourhood Filter," AEU - International Journal of Electronics and Communications, vol. 62, no. 6, pp. 472-477, Jun. 2008. DOI: 10.1016/j.aeue.2007.07.001.
  5. X. Xiao, N. N. Xiong, J. Lai, C. D. Wang, Z. Sun, and J. Yan, "A Local Consensus Index Scheme for Random-Valued Impulse Noise Detection Systems," IEEE Transactions on Systems, Man, and Cybernetics: Systems, vol. 51, no. 6, pp. 3412-3428, Jun. 2021. DOI: 10.1109/TSMC.2019.2925886.
  6. P. Goyal and V. Chaurasia, "Application of Median Filter in Removal of Random Valued Impulse Noise from Natural Images," in Proceedings of 2017 International conference of Electronics, Communication and Aerospace Technology (ICECA), Coimbatore, India, pp. 125-128, 2017. DOI: 10.1109/ICECA.2017.8203657.
  7. B. W. Cheon and N. H. Kim, "A Modified Steering Kernel Filter for AWGN Removal based on Kernel Similarity," Journal of information and communication convergence engineering, vol. 20, no. 3, pp. 195-203, Sep. 2022. DOI: 10.56977/jicce.2022.20.3.195.
  8. B. W. Cheon and N. H. Kim, "Modified Gaussian Filter based on Fuzzy Membership Function for AWGN Removal in Digital Images," Journal of information and communication convergence engineering, vol. 19, no. 1, pp. 54-60, Mar. 2021. DOI: 10.6109/jicce.2021.19.1.54.