Non-uniform motion deblurring with blurry component divided guidance

Wang, Pei; Sun, Wei; Yan, Qingsen; Niu, Axi; Li, Rui; Zhu, Yu; Sun, Jinqiu; Zhang, Yanning

doi:10.1016/j.patcog.2021.108082

Cited by 18 publications

(5 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Most existing learning-based deblurring methods [36,37] usually perform deblurring in a single spatial domain. If the training datasets are not realistic enough, the deblurring models trained on them may produce resultant images with unpleasant artifacts.…”

Section: Spatial Fourier Reconstruction Blockmentioning

confidence: 99%

Real‐world image deblurring using data synthesis and feature complementary network

Wei,

Ge,

Qiao

et al. 2024

IET Image Processing

View full text Add to dashboard Cite

Many learning‐based approaches to image deblurring have received increasing attention in recent years. However, the models trained on existing synthetic datasets do not generalize well to real‐world blur, resulting in undesirable artifacts and residual blur. This work attempts to address this problem from two aspects: training data synthesis and network architecture. To narrow the domain gap between synthetic and real domains, a realistic blur synthesis pipeline to generate high‐quality blurred data is proposed. Since the blur is non‐uniform and has different scales and degrees, a parallel feature complementary module to fully exploit the local and non‐local information, which improves the feature representation and helps the network to perceive the non‐uniform blur, is developed. In addition, a spatial Fourier reconstruction block to facilitate correct detail recovery in the spatial and Fourier domains is introduced. Based on these two designs, an effective encoder–decoder network for deblurring is designed. Extensive experiments demonstrate the validity and superiority of the proposed blur synthesis method and deblurring network. In particular, the proposed deblurring network can achieve superior or comparable performance to Restormer, while saving 70% of network parameters and 53% of floating point operations (FLOPs).

show abstract

Section: Spatial Fourier Reconstruction Blockmentioning

confidence: 99%

Real‐world image deblurring using data synthesis and feature complementary network

Wei,

Ge,

Qiao

et al. 2024

IET Image Processing

View full text Add to dashboard Cite

show abstract

“…Deep learning techniques were also extended to solve the motion blur problem in radar images. [41] highlighted the challenge of non-uniform motion blur in radar images and proposed a deep learning-based approach for deblurring. The method incorporates blurry component separation to handle complex motion blur patterns.…”

Section: Related Workmentioning

confidence: 99%

Single Radar Image Motion Deblurring Through PSF Estimation and Convolution Regularizer

Sharma,

García-Fernández,

Álvarez-Narciandi

et al. 2024

Trans. Rad. Sys.

View full text Add to dashboard Cite

This paper proposes a deep learning-based motion deblurring method for radar images using point-spread function (PSF) estimations. Motion blur is a common problem in radar imaging due to the movement of the radar platform or the target. The proposed method uses a convolutional neural network (CNN) to learn the mapping between the blurred and sharp images. In addition, the PSF of the motion blur is estimated using a separate CNN. The estimated point spread function (PSF) is utilized in conjunction with the input image to reconstruct a deblurred image. The reconstruction process is further optimized by incorporating the relationship between the input image, PSF, and the ground truth into the training loss term and optimizing it. The network is trained on a large dataset of simulated blurred and sharp radar images with corresponding PSFs. The developed method is evaluated on blurred radar images, subjected to varying degrees and lengths of blur, and compared with state-of-the art methods. The results show that the proposed method outperforms the existing methods both qualitatively and quantitatively. The proposed method can be used for various radar imaging applications, such as target detection and recognition, surveillance and navigation.

show abstract

“…In dynamic scenes, when the camera exposure setting is set long to ensure sufficient light, motions will be apparent and deblurring methods are required. Representative models include conventional approaches [33,36] and CNN based models [7,13,19,24,29,40,42]. Meanwhile, when exposure is set short to avoid the superposition of motions, noise artifacts will be obvious due to insufficient light, and denoising approaches are needed, including conventional techniques [2,9,14,35] and CNN based methods [3,15,40,41,43].…”

Section: Related Work 21 Image Restorationmentioning

confidence: 99%

Exploring and Evaluating Image Restoration Potential in Dynamic Scenes

Zhang¹,

Su²,

Zhu³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

In dynamic scenes, images often suffer from dynamic blur due to superposition of motions or low signal-noise ratio resulted from quick shutter speed when avoiding motions. Recovering sharp and clean results from the captured images heavily depends on the ability of restoration methods and the quality of the input. Although existing research on image restoration focuses on developing models for obtaining better restored results, fewer have studied to evaluate how and which input image leads to superior restored quality. In this paper, to better study an image's potential value that can be explored for restoration, we propose a novel concept, referring to image restoration potential (IRP). Specifically, We first establish a dynamic scene imaging dataset containing composite distortions and applied image restoration processes to validate the rationality of the existence to IRP. Based on this dataset, we investigate several properties of IRP and propose a novel deep model to accurately predict IRP values. By gradually distilling and selective fusing the degradation features, the proposed model shows its superiority in IRP prediction. Thanks to the proposed model, we are then able to validate how various image restoration related applications are benefited from IRP prediction. We show the potential usages of IRP as a filtering principle to select valuable frames, an auxiliary guidance to improve restoration models, and even an indicator to optimize camera settings for capturing better images under dynamic scenarios.

show abstract

Non-uniform motion deblurring with blurry component divided guidance

Cited by 18 publications

References 14 publications

Real‐world image deblurring using data synthesis and feature complementary network

Real‐world image deblurring using data synthesis and feature complementary network

Single Radar Image Motion Deblurring Through PSF Estimation and Convolution Regularizer

Exploring and Evaluating Image Restoration Potential in Dynamic Scenes

Contact Info

Product

Resources

About