Snow Removal From Light Field Images

Yan, Tao; Ding, Yu-Yang; Zhang, Fan; Xie, Ningyu; Liu, Wenxi; Wu, Zhengtian; Liu, Yuan

doi:10.1109/access.2019.2951917

Cited by 9 publications

(4 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…One of the typical problems of such characteristics is image denoising, especially impulsive noise suppression, which has been recently explored and become a popular topic to examine [ 7 , 8 , 9 , 10 , 11 , 12 , 13 ]. Other applications, where larger artifacts are detected and removed from the images but not all the pixels are altered, include rain [ 14 , 15 , 16 , 17 , 18 , 19 , 20 ], snow [ 20 , 21 , 22 , 23 , 24 ], marine snow [ 25 ], and crack [ 26 ] removal problems. Furthermore, algorithms that are dedicated to image tamper detection and correction [ 27 ] may fit into the above-mentioned description.…”

Section: Introductionmentioning

confidence: 99%

Decomposed Dissimilarity Measure for Evaluation of Digital Image Denoising

Maliński

2023

Sensors

View full text Add to dashboard Cite

A new approach to the evaluation of digital image denoising algorithms is presented. In the proposed method, the mean absolute error (MAE) is decomposed into three components that reflect the different cases of denoising imperfections. Moreover, aim plots are described, which are designed to be a very clear and intuitive form of presentation of the new decomposed measure. Finally, examples of the application of the decomposed MAE and the aim plots in the evaluation of impulsive noise removal algorithms are presented. The decomposed MAE measure is a hybrid of the image dissimilarity measure and detection performance measures. It provides information about sources of errors such as pixel estimation errors, unnecessary altered pixels, or undetected and uncorrected distorted pixels. It measures the impact of these factors on the overall correction performance. The decomposed MAE is suitable for the evaluation of algorithms that perform a detection of the distortion that affects only a certain fraction of the image pixels.

show abstract

Section: Introductionmentioning

confidence: 99%

Decomposed Dissimilarity Measure for Evaluation of Digital Image Denoising

Maliński

2023

Sensors

View full text Add to dashboard Cite

show abstract

“…Unlike the previous GAN, their generator network comprises a clean background module and a snow mask estimation module to extract useful information. Based on a 3D residual network, Yan et al [15] utilized both contextual information and 3D scene structure information to effectively detect snowflakes of different sizes in low frequency (LF) images. Finally, an encoder-decoder-based LF image restoration network was proposed to restore the background image.…”

Section: Introductionmentioning

confidence: 99%

Single-Image Snow Removal Based on an Attention Mechanism and a Generative Adversarial Network

Jia

Yang

et al. 2021

IEEE Access

View full text Add to dashboard Cite

Bad weather, such as snowfall, can seriously decrease the quality of images and pose great challenges to computer vision algorithms. In view of the negative effect of snowfall, this paper presents a single-image snow removal method based on a generative adversarial network (GAN). Unlike previous GANs, our GAN includes an attention mechanism in the generator component. By injecting attention information, the network can pay increased attention to areas covered by snow and improve its capability to perform local repairs. At the same time, we improve the traditional U-Net network by combining it with the residual network to enhance the effect of the model when removing snowflakes from a single image. Our experiments on both synthetic and real-word images show that our method produces better results than those of other state-of-the-art methods.INDEX TERMS Snow removal, generative adversarial networks, attention mechanisms.

show abstract

“…Several non-deep-learning-based rain removal methods, such as frequency domain representation [10], Gaussian mixture model [11], and sparse representation [12], have been proposed and demonstrated to lead to significant quality improvements. Owing to increasing interest in deep learning recently, few deep-learning-based rain removal methods have been proposed [1]- [5], [13]- [14], [27]- [32], [49], [50]- [57]. Deep networks allow us to easily learn the correlation between rain streaks and background and typically achieve better performance than non-deep-learning approaches.…”

Section: Introductionmentioning

confidence: 99%