Temat: dehazing - Katalog OPAC zbiorów

Skocz do pozycji: 1.

Tytuł:: A fast dehazing method with high color fidelity
Autorzy:: Shan, Y.
Ge, P.
Powiązania:: https://bibliotekanauki.pl/articles/174419.pdf
Data publikacji:: 2018
Wydawca:: Politechnika Wrocławska. Oficyna Wydawnicza Politechniki Wrocławskiej
Tematy:: digital image processing
atmospheric scattering
dehazing
Opis:: In this paper, we present a fog image degradation model with the combination of HSI color space and the atmosphere scattering model. Based on the model, a fast dehazing method with high color fidelity has been proposed. The main advantage of the proposed method compared with others is its speed. This speed allows the method to be applied within real-time processing applications as a step of preprocessing. Another advantage is the possibility to handle both color images and gray level images. The algorithm depends only on on two parameters, and both are easy to set. Experiments on haze images demonstrate that the proposed method can achieve wonderful image visibility with a higher computing speed.
Źródło:: Optica Applicata; 2018, 48, 3; 515-527
0078-5466
1899-7015
Pojawia się w:: Optica Applicata
Dostawca treści:: Biblioteka Nauki

Artykuł

Skocz do pozycji: 2.

Tytuł:: Underwater image enhancement via efficient generative adversarial network
Autorzy:: Qian, Xin
Ge, Peng
Powiązania:: https://bibliotekanauki.pl/articles/2033886.pdf
Data publikacji:: 2021
Wydawca:: Politechnika Wrocławska. Oficyna Wydawnicza Politechniki Wrocławskiej
Tematy:: underwater
image dehazing
generative adversarial network
GAN
Opis:: Underwater image enhancement has been receiving much attention due to its significance in facilitating various marine explorations. Inspired by the generative adversarial network (GAN) and residual network (ResNet) in many vision tasks, we propose a simplified designed ResNet model based on GAN called efficient GAN (EGAN) for underwater image enhancement. In particular, for the generator of EGAN we design a new pair of convolutional kernel size for the residual block in the ResNet. Secondly, we abandon batch normalization (BN) after every convolution layer for faster training and less artifacts. Finally, a smooth loss function is introduced for halo-effect alleviation. Extensive qualitative and quantitative experiments show that our methods accomplish considerable improvements compared to the state-of-the-art methods.
Źródło:: Optica Applicata; 2021, 51, 4; 483-497
0078-5466
1899-7015
Pojawia się w:: Optica Applicata
Dostawca treści:: Biblioteka Nauki

Artykuł

Informacja