분해 심층 학습을 이용한 저조도 영상 개선 방식
Low-light Image Enhancement Method Using Decomposition-based Deep-Learning
1
조회
0
다운로드
1
인용
0
참고문헌
키워드
초록
본 논문에서는 저조도 영상을 개선하기 위한 영상 분해 기반 심층 학습 방법 및 분해 채널 특성에 따른 손실함수를 제안한다. 기존 기법들의 문제점인 색신호 왜곡 및 할로 현상을 제거하기 위해, 입력 영상의 휘도 채널을 반사 성분과 조도 성분으로 분해하고, 반사 성분, 조도 성분 및 색차 신호를 신호 특성에 적합한 심층학습 과정을 적용하는 분해 기반 다중 구조 심층학습 방법을 제안한다. 더불어, 분해 채널들의 특성에 따른 혼합 놈 기반의 손실함수를 정의하여 복원 영상의 안정성을 증대하고 열화 현상을 제거하기 위한 기법에 대해 기술한다. 실험 결과를 통해 제안한 방법이 다양한 저조도 영상을 효과적으로개선하였음을 확인할 수 있었다.
Abstract
This paper introduces an image decomposition-based deep learning method and loss function to improve low-lightimages. In order to remove color distortion and halo artifact, illuminance channel of an input image is decomposed intoreflectance and luminance channels, and a decomposition-based multiple structural deep learning process is applied toeach channel. In addition, a mixed norm-based loss function is described to increase the stability and remove blurring inreconstructed image. Experimental results show that the proposed method effectively improve various low-light images.
참고문헌 (25)
- Wencheng Wang, "An Experiment-Based Review of Low-Light Image Enhancement Methods", IEEE Access, 8, pp.87884~87917 (2020), Institute of Electrical and Electronics Engineers (IEEE)[DOI]
- E. H. Land, "Lightness and retinex theory", J. of Opt. Soc. of America, 61(1), pp.1~11 (1971)[DOI]
- D. J. Jobson, "Properties and performance of a center", surround retinex, 1997, pp.6(3), 451~461
- D. J. Jobson, "A multi-scale retinex for bridging the gap between color images and the human observation of scenes", IEEE Trans. on Image Process, 6(7), pp.965~976[DOI]
- E. Provenzi, "Random spray retinex:A new retinex implementation to investigate the local properties of the model", IEEE Trans. on Image Process, 16(1), pp.162~171 (2007)[DOI]
- T. Celik, "Spatial entropy-based global and local image contrast enhancement", IEEE Trans. on Image Process, 23(12), pp.5298~5308 (2014)[DOI]
- Y. Shin, "Efficient naturalness restoration for non-uniform illumination images", IET Image Process, 9(9), pp.662~671 (2015)
- M. Lecca, "GRASS:A gradient-based random sampling scheme for Milano retinex", IEEE Trans. on Image Process, 26(6), pp.2767~2780 (2017)[DOI]
- R. Kimmel, "A variational framework for retinex", Int. J. of Comput. Vis, 52, pp.7~23 (2003)[DOI]
- D. Zosso, "Non-local retinex-A unifying framework and beyond", SIAM J. on Image Sc, 8(2), pp.787~826 (2015)[DOI]
- S. Park, "Low-light image enhancement using variational optimization-based retinex model", IEEE Trans. on Consumer Electron, 63(2), pp.178~184 (2017)[DOI]
- K. G. Lore, "LLNet: A deep autoencoder approach to natural low-light image enhancement", Pattern Recognition, 61, pp.650~662 (2017)[DOI]
- L. Shen, "Msr-net: lowlight image enhancement using deep convolutional network" (2017)
- X. Guo, "Lime: Low-light image enhancement via illumination map estimation", IEEE Trans. on Image Process, 26(2), pp.982~999 (2017)[DOI]
- "2018", ~
- F. Lv, "MBLLEn:Low-light image", video enhancement using CNNs, 2019, pp.1(13), 1~13
- Y. Zhang, "Kindling the darkness: A practical low-light image enhancer", Proc. of ACM Int. Conf. on Mult, 1632(1640), pp.1632~1640 (2019)
- J. G. Oh, "Adaptive image rendering using a nonlinear mapping function based retinex model", Sensors, 19(4), pp.969~ (2019)[DOI]
- D. P. Kingma, "Adam: A method for stochastic optimization" (2017)
- "Live image quality assessment database release 2"[URL]
- "ImageNet"[URL]
- "Electromagnetics and Sensors Research Branch"[URL]
- "The Berkeley Segmentation Dataset and Benchmark"[URL]
- Shuhang Wang, "Naturalness Preserved Enhancement Algorithm for Non-Uniform Illumination Images", IEEE Transactions on Image Processing, 22(9), pp.3538~3548 (2013), Institute of Electrical and Electronics Engineers (IEEE)[DOI]
- A. Mittal, "Making a “Completely Blind” Image Quality Analyzer", IEEE Signal Processing Letters, 20(3), pp.209~212 (2013), Institute of Electrical and Electronics Engineers (IEEE)[DOI]