Image Shadow Removal Using End-To-End Deep Convolutional Neural Networks

Fan, Hui; Han, Meng; Li, Jinjiang

doi:10.3390/app9051009

Cited by 21 publications

(10 citation statements)

References 40 publications

(92 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…According to previous observations [2,26,27], a shadow image I s can be generated from the pixelwise product of a shadow-free image I sf and a shadow matte α, as shown in (1).…”

Section: Methodsmentioning

confidence: 99%

C2ShadowGAN: cycle-in-cycle generative adversarial network for shadow removal using unpaired data

et al. 2022

View full text Add to dashboard Cite

Recent advances in deep learning technology, and the availability of public shadow image datasets, have enabled significant performance improvements of shadow removal tasks in computer vision. However, most deep learning-based shadow removal methods are usually trained in a supervised manner, in which paired shadow and shadow-free data are required. We developed a weakly supervised generative adversarial network with a cycle-in-cycle structure for shadow removal using unpaired data. In addition, we introduced new loss functions to reduce unnecessary transformations for non-shadow areas and to enable smooth transformations for shadow boundary areas. We conducted extensive experiments using the ISTD and Video Shadow Removal datasets to assess the effectiveness of our methods. The experimental results show that our method is superior to other state-of-the-art methods trained on unpaired data.

show abstract

“…According to previous observations [2,26,27], a shadow image I s can be generated from the pixelwise product of a shadow-free image I sf and a shadow matte α, as shown in (1).…”

Section: Methodsmentioning

confidence: 99%

C2ShadowGAN: cycle-in-cycle generative adversarial network for shadow removal using unpaired data

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Thus, it is different from other intelligent algorithm tools. Among the networks, the most widely used is BPNN [23,24]. The composition of the BPNN model includes the input layer, the hidden layer, and the output layer.…”

Section: Bpnn Algorithm Based On Deep Learningmentioning

confidence: 99%

The Analysis of the Urban Sprawl Measurement System of the Yangtze River Economic Belt, Based on Deep Learning and Neural Network Algorithm

Huafang

Cheng

2020

IJERPH

View full text Add to dashboard Cite

In the context of rapid urbanization, the spread of cities in the Yangtze River Economic Belt is intensifying, which has an impact on the green and sustainable development of these cities. It is necessary to establish an accurate urban sprawl measurement system. First, the regulation theory of urban sprawl is explained. According to the actual development situation of cities in the Yangtze River Economic Belt, smart growth theory is selected as the basic regulation method of urban sprawl. Second, the back propagation neural network (BPNN) algorithm under deep supervised learning is applied to construct a smart evaluation model of land use growth. Finally, based on the actual development of cities in the Yangtze River Economic Belt, the quantitative growth measurement method is selected to construct a measurement system of urban sprawl in the Yangtze River Economic Belt, and the empirical analysis is carried out. The training results show that the proposed BPNN smart growth evaluation model, based on deep supervised learning, has good evaluation accuracy, and the error is within the preset range. The analysis of the quantitative growth-based measurement system in the increase of urban construction land shows that the increase in urban construction land area of the Yangtze River Economic Belt from 2014 to 2019 was 78.67 km2. Meanwhile, the increases in urban construction land area in different years are different. The empirical results show that the population composition of the Yangtze River Economic Belt and the urban construction area between 2005 and 2019 show a trend of increasing annually; at the same time, urban sprawl development shows a staged characteristic. It is of great significance to apply deep learning fusion neural network algorithm in the construction of the urban sprawl measurement system, which provides a quantitative basis for the in-depth analysis and discussion of urban sprawl.

show abstract

“…A comprehensive detail of the various shadow detection and removal methods is presented by Murali et al [ 21 ]. More recently, ANN-based deep learning methods have also been deployed for shadow detection and removal [ 22 , 23 , 24 ]. Although several researchers have tried to solve the problem of shadow removal; this task still remains a complex topic with moderate to good success [ 21 ].…”

Section: Introductionmentioning

confidence: 99%

Automatic Detection of Cracks on Concrete Surfaces in the Presence of Shadows

Palevicius

Pal

Landauskas

et al. 2022

Sensors

View full text Add to dashboard Cite

Deep learning-based methods, especially convolutional neural networks, have been developed to automatically process the images of concrete surfaces for crack identification tasks. Although deep learning-based methods claim very high accuracy, they often ignore the complexity of the image collection process. Real-world images are often impacted by complex illumination conditions, shadows, the randomness of crack shapes and sizes, blemishes, and concrete spall. Published literature and available shadow databases are oriented towards images taken in laboratory conditions. In this paper, we explore the complexity of image classification for concrete crack detection in the presence of demanding illumination conditions. Challenges associated with the application of deep learning-based methods for detecting concrete cracks in the presence of shadows are elaborated on in this paper. Novel shadow augmentation techniques are developed to increase the accuracy of automatic detection of concrete cracks.

show abstract

Image Shadow Removal Using End-To-End Deep Convolutional Neural Networks

Cited by 21 publications

References 40 publications

C2ShadowGAN: cycle-in-cycle generative adversarial network for shadow removal using unpaired data

C2ShadowGAN: cycle-in-cycle generative adversarial network for shadow removal using unpaired data

The Analysis of the Urban Sprawl Measurement System of the Yangtze River Economic Belt, Based on Deep Learning and Neural Network Algorithm

Automatic Detection of Cracks on Concrete Surfaces in the Presence of Shadows

Contact Info

Product

Resources

About