Improving Performance of Machine Learning-based Haze Removal Algorithms with Enhanced Training Database
0
조회
0
다운로드
2
인용
0
참고문헌
키워드
초록
Haze removal is an object of scientific desire due to its various practical applications. Existing algorithms arefounded upon histogram equalization, contrast maximization, or the growing trend of applying machine learning inimage processing. Since machine learning-based algorithms solve problems based on the data, they usuallyperform better than those based on traditional image processing/computer vision techniques. However, to achievesuch a high performance, one of the requisites is a large and reliable training database, which seems to beunattainable owing to the complexity of real hazy and haze-free images acquisition. As a result, researchers arecurrently using the synthetic database, obtained by introducing the synthetic haze drawn from the standarduniform distribution into the clear images. In this paper, we propose the enhanced equidistribution, improving uponour previous study on equidistribution, and use it to make a new database for training machine learning-basedhaze removal algorithms. A large number of experiments verify the effectiveness of our proposed methodology
Abstract
Haze removal is an object of scientific desire due to its various practical applications. Existing algorithms arefounded upon histogram equalization, contrast maximization, or the growing trend of applying machine learning inimage processing. Since machine learning-based algorithms solve problems based on the data, they usuallyperform better than those based on traditional image processing/computer vision techniques. However, to achievesuch a high performance, one of the requisites is a large and reliable training database, which seems to beunattainable owing to the complexity of real hazy and haze-free images acquisition. As a result, researchers arecurrently using the synthetic database, obtained by introducing the synthetic haze drawn from the standarduniform distribution into the clear images. In this paper, we propose the enhanced equidistribution, improving uponour previous study on equidistribution, and use it to make a new database for training machine learning-basedhaze removal algorithms. A large number of experiments verify the effectiveness of our proposed methodology
참고문헌 (8)
- Z. Xu, "Fog Removal from Color Images using Contrast Limited Adaptive Histogram Equalization", 2009 2nd International Congress on Image and Signal Processing, pp.1~5 (2009)
- R. Tan, "Visibility in bad weather from a single image", 2008 IEEE Conference on Computer Vision and Pattern Recognition, pp.1~8 (2008)
- Qingsong Zhu, "A Fast Single Image Haze Removal Algorithm Using Color Attenuation Prior", IEEE Transactions on Image Processing, 24(11), pp.3522~3533 (2015), Institute of Electrical and Electronics Engineers (IEEE)[DOI]
- D. Ngo, "A New Data Preparation Methodology in Machine Learning-based Haze Removal Algorithms", International Conference on Electronics, Information, and Communication (ICEIC) 2019 (2019)
- K. Ma, "Perceptual evaluation of single image dehazing algorithm", 2015 IEEE International Conference on Image Processing (ICIP), pp.3600~3604 (2015)
- J-P Tarel, "Vision Enhancement in Homogeneous and Heterogeneous Fog", IEEE Intelligent Transportation Systems Magazine, 4(2), pp.6~20 (2012), Institute of Electrical and Electronics Engineers (IEEE)[DOI]
- Nicolas Hautière, "BLIND CONTRAST ENHANCEMENT ASSESSMENT BY GRADIENT RATIOING AT VISIBLE EDGES", Image Analysis & Stereology, 27(2), pp.87~95 (2008), Slovenian Society for Stereology and Quantitative Image Analysis[DOI]
- Hojatollah Yeganeh, "Objective Quality Assessment of Tone-Mapped Images", IEEE Transactions on Image Processing, 22(2), pp.657~667 (2013), Institute of Electrical and Electronics Engineers (IEEE)[DOI]