Blind inpainting using the fully convolutional neural network

Cai, Nian; Su, Zhenghang; Lin, Zhineng; Wang, Han; Yang, Zhijing; Ling, Bingo Wing‐Kuen

doi:10.1007/s00371-015-1190-z

Cited by 88 publications

(37 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There also exists texture‐based inpainting approaches which attempt to preserve the texture of an image throughout the inpainting domain; an example of texture‐based inpainting can be found in the study by Cai et al In addition, there have been advances in machine learning algorithms to determine inpainting domains. For instance, blind inpainting determines which regions to inpaint based on their different texture from the background . Such an algorithm could allow for automatic detection of all RBC regions in the tissue.…”

Section: Discussionmentioning

confidence: 99%

Using digital inpainting to estimate incident light intensity for the calculation of red blood cell oxygen saturation from microscopy images

Sové

Drakos

Fraser

et al. 2018

Journal of Biophotonics

View full text Add to dashboard Cite

Red blood cell oxygen saturation (SO ) is an important indicator of oxygen supply to tissues in the body. SO can be measured by taking advantage of spectroscopic properties of hemoglobin. When this technique is applied to transmission microscopy, the calculation of saturation requires determination of incident light intensity at each pixel occupied by the red blood cell; this value is often approximated from a sequence of images as the maximum intensity over time. This method often fails when the red blood cells are moving too slowly, or if hematocrit is too large since there is not a large enough gap between the cells to accurately calculate the incident intensity value. A new method of approximating incident light intensity is proposed using digital inpainting. This novel approach estimates incident light intensity with an average percent error of approximately 3%, which exceeds the accuracy of the maximum intensity-based method in most cases. The error in incident light intensity corresponds to a maximum error of approximately 2% saturation. Therefore, though this new method is computationally more demanding than the traditional technique, it can be used in cases where the maximum intensity-based method fails (eg, stationary cells), or when higher accuracy is required.

show abstract

Section: Discussionmentioning

confidence: 99%

Using digital inpainting to estimate incident light intensity for the calculation of red blood cell oxygen saturation from microscopy images

Sové

Drakos

Fraser

et al. 2018

Journal of Biophotonics

View full text Add to dashboard Cite

show abstract

“…The rise of neural networks has inspired the idea of using neural networks to solve real-world problems. In [11][12][13][14], everyone discusses the self-training and learning of the image to be repaired through the introduction of the neural network and extracts the features for completion, each of whom has obtained a relatively ideal effect. A novel network based on sparse coding and the noise reduction autoencoder deep network is put forward in [15], which has achieved the removal of the image overlay text and the repair of deleted areas.…”

Section: B Neural Network Methodsmentioning

confidence: 99%

3D Model Inpainting Based on 3D Deep Convolutional Generative Adversarial Network

Wang

2020

IEEE Access

View full text Add to dashboard Cite

In recent years, the problem of hole repairing in the 3D model has been widely concerned in related fields. As the Generative Adversarial Network (GAN) has achieved great success in generating realistic images, a 3D mesh model repair method based on the 3D Deep Convolutional Generative Adversarial Network (3D-DCGAN) is proposed in this paper. The algorithm contains two GANs: a local GAN and a global GAN. Four steps have been used to implement this concept. First, the 3D model is voxelized, and a mask is used to identify the repairing area; Second, the repairing area is generated by training local GAN; Third, the repaired region is combined with the 3D model to be repaired, thereafter, the global GAN is trained with the combined model. Finally, a decent repaired model is obtained with the perfect transition. The experimental results show that this algorithm can effectively generate the repairing area while retaining the details of the area and blend it with the model to be repaired. INDEX TERMS 3D model inpainting, GAN, local GAN, global GAN.

show abstract

“…The class of techniques designed to replace empty regions in an image with perceptually plausible content is called inpainting after Bertalmió et al [Ber00a]. Inpainting can be used for many purposes in visual computing, including, for example, denoising [Ad17a], image compression [Mai09], or automatic repair of damaged images [Cai17a]. There are many technical approaches to inpainting, cf.…”

Section: Introductionmentioning

confidence: 99%

Morphological Amoeba-based Patches for Exemplar-Based Inpainting

Castillo

Cunningham

Winger³

et al. 2018

JWSCG

View full text Add to dashboard Cite

Inpainting is the process of replacing areas in an image with a perceptually plausible substitution. A common technique is to iteratively match and fill small patches at the edge of the target region making use of similar patches from the same image. Nearly all inpainting algorithms based on this approach use a single patch size for the entire image. Yet, it seems clear that differently sized structures within the same image-for example a leaf versus a car tire-may require different patch sizes in order to achieve reasonable inpainting results. Likewise, a fixed patch size will give different results for the same image when the image resolution is doubled. A reasonable patch should therefore take into account the overall image size as well as the size and shape of the structures at the patch location. The aim of our paper is to study the effect of adaptively altering size and shape of the patch. We show that this technique leads to a better quality of the inpainting result compared to a fixed patch size.

show abstract

Blind inpainting using the fully convolutional neural network

Cited by 88 publications

References 17 publications

Using digital inpainting to estimate incident light intensity for the calculation of red blood cell oxygen saturation from microscopy images

Using digital inpainting to estimate incident light intensity for the calculation of red blood cell oxygen saturation from microscopy images

3D Model Inpainting Based on 3D Deep Convolutional Generative Adversarial Network

Morphological Amoeba-based Patches for Exemplar-Based Inpainting

Contact Info

Product

Resources

About