Image Demosaicking Using Densely Connected Convolutional Neural Network

  • Park, Bumjun
  • Jeong, Je chang
Citations

SCOPUS

2

초록

In this paper, we propose image demosaicking model using densely connected convolutional neural network. Recently, deep neural networks show improved results in image processing field compared with conventional algorithms. However, they often suffer vanishing-gradient problem which makes models hard to be trained. To solve this problem, we applied densely connected convolutional neural network. More than that, our proposed network doesn't need any initial interpolation which can reduce computational complexity. Finally, we applied sub-pixel interpolation layer which can generate demosaicked output image efficiently and accurately. Experimental results show that our proposed model outperformed conventional methods.

키워드

Color filter array interpolationConvolutional neural networkDeep learningDemosaickingConvolutionDeep learningImage enhancementInterpolationNeural networksColor filter array interpolationConventional algorithmsConventional methodsConvolutional neural networkDemosaickingSubpixel interpolationVanishing gradientDeep neural networks
제목
Image Demosaicking Using Densely Connected Convolutional Neural Network
저자
Park, BumjunJeong, Je chang
DOI
10.1109/SITIS.2018.00053
발행일
2018-07
유형
Conference Paper
저널명
Proceedings - 14th International Conference on Signal Image Technology and Internet Based Systems, SITIS 2018
페이지
304 ~ 307