Journal of IKEEE (전기전자학회논문지)

Institute of Korean Electrical and Electronics Engineers (한국전기전자학회)
권호 표지
DOI QR코드

DOI QR Code

Implementation of Deep CNN denoiser for Reducing Over blur

Over blur를 감소시킨 Deep CNN 구현

  • Lee, Sung-Hun (Dept. of Computer Engineering, Seokyeong University) ;
  • Lee, Kwang-Yeob (Dept. of Computer Engineering, Seokyeong University) ;
  • Jung, Jun-Mo (Dept. of Electronics Engineering, Seokyeong University)
  • Received : 2018.12.12
  • Accepted : 2018.12.17
  • Published : 2018.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

In this paper, we have implemented a network that overcomes the over-blurring phenomenon that occurs when removing Gaussian noise. In the conventional filtering method, blurring of the original image is performed to remove noise, thereby eliminating high frequency components such as edges and corners. We propose a network that reducing over blurring while maintaining denoising performance by adding denoised high frequency components to denoisers based on CNN.

본 논문에서, Gaussian noise를 제거할 때 발생하는 over blurring 현상을 감소시키는 network를 구현하였다. 기존 filtering 방식은 원 영상을 blurring하여 noise를 제거함으로써, edge나 corner 같은 high frequency 성분도 함께 지워지는 것을 확인할 수 있다. CNN (Convolutional Neural Network)기반 denoiser의 경우도 사소한 edge, keypoint를 noise로 인식하여 이러한 정보를 잃게 된다. 우리는 CNN을 기반으로 denoising된 high frequency 성분만을 획득하여 기존 denoiser에 추가함으로써 denoising 성능을 유지하면서 over blurring을 완화하는 network 제안한다.

Keywords

JGGJB@_2018_v22n4_1242_f0001.png 이미지

Fig. 1. DnCNN’s architecture. 그림 1. DnCNN 구조

JGGJB@_2018_v22n4_1242_f0002.png 이미지

Fig. 2. Proposed network’s architecture. 그림 2. 제안하는 network 구조

JGGJB@_2018_v22n4_1242_f0003.png 이미지

Fig. 3. ResNet team’s full pre-activation. 그림 3. ResNet team의 full pre-activation

JGGJB@_2018_v22n4_1242_f0004.png 이미지

Fig. 4. Low frequency component. 그림 4. Low frequency 성분

JGGJB@_2018_v22n4_1242_f0005.png 이미지

Fig. 5. Overall data flow. 그림 5. 전반적인 데이터 흐름

JGGJB@_2018_v22n4_1242_f0006.png 이미지

Fig. 6. DnCNN’s denoising result. 그림 6. DnCNN의 denoising 결과

JGGJB@_2018_v22n4_1242_f0007.png 이미지

Fig. 7. Proposed network’s result. 그림 7. 제안하는 network의 결과

Table 1. The PSNR according to subtraction. 표 1. 감산 여부에 따른 PSNR

JGGJB@_2018_v22n4_1242_t0001.png 이미지

Table 2. The PSNR of each training method. 표 2. 학습방법에 따른 PSNR 결과

JGGJB@_2018_v22n4_1242_t0002.png 이미지

References

  1. Jeon, Yougn-Eun, et al. "Gaussian Noise Reduction Algorithm using Self-similarity," Journal of the Institute of Electronics Engineers of Korea SP, 44, 5, 1-10, 2007.
  2. Multimedia Laboratory, "Large-scale CelebFaces Attributes (CelebA) Dataset," https://mmlab.ie.cuhk.edu.hk
  3. Zhang, K., Zuo, W., Chen, Y., Meng, D., & Zhang, L, "Beyond a gaussian denoiser: Residual learning of deep cnn for image denoising," IEEE Transactions on Image Processing, 26, 7, 3142-3155, 2017. DOI:10.1109/TIP.2017.2662206
  4. He, K., Zhang, X., Ren, S., & Sun, J, "Deep residual learning for image recognition," In Proceedings of the IEEE conference on computer vision and pattern recognition. 770-778, 2016.
  5. He, K., Zhang, X., Ren, S., & Sun, J., "Identity mappings in deep residual networks," In European conference on computer vision, (2016, October), 630-645.