FDnCNN-based image denoising for multi-labfel localization measurement

Li, Lin; Yu, Xiaolei; Jin, Zhixin; Zhao, Zhimin; Zhuang, Xiao; Liu, Zhenlu

doi:10.1016/j.measurement.2019.107367

Cited by 14 publications

(9 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While discriminative learning methods have fast testing speed but are limited within a specific task. FD-nCNN [17] introduces a module named fusion block in CNNs to obtain high-quality images in real-time. Inheriting from DIP, Noise2Noise [16] employs pairs of noisy images with the same content for supervised learning, while Noise2Void [13] is a self-supervised training method with any noisy images.…”

Section: Related Workmentioning

confidence: 99%

“…Although several methods have successfully incorporated into image denoising, e.g. (C)DnCNN [44], FDnCNN [17] and Noise2Self [3], they are usually time-consuming, thus it still lacks of the computationally efficient denoising solution in the literature. In image denoising, given a clear image I P R W ˆHˆL , the additive noise-corrupted image…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

IDEA-Net: Adaptive Dual Self-Attention Network for Single Image Denoising

Zuo

Chen²,

Han

et al. 2022

2022 IEEE/CVF Winter Conference on Applications of Computer Vision Workshops (WACVW)

View full text Add to dashboard Cite

Image denoising is a challenging task due to possible data bias and prediction variance. Existing approaches usually suffer from high computational cost. In this work, we propose an unsupervised image denoiser, dubbed as adaptIve Dual sElf-Attention Network (IDEA-Net), to handle these challenges. IDEA-Net benefits from a generatively learned image-wise dual self-attention region where the denoising process is enforced. Besides, IDEA-Net is not only robust to possible data bias but also helpful to reduce the prediction variance by applying a simplified encoder-decoder with Poisson dropout operations on a single noisy image merely. The proposed IDEA-Net demonstrated the outperformance on four benchmark datasets compared with other single-image-based learning and nonlearning image denoisers. IDEA-Net also shows an appropriate choice to remove real-world noise in low-light and noisy scenes, which in turn, contribute to more accurate dark face detection. The source code is available at https://github.com/zhemingzuo/IDEA-Net.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

IDEA-Net: Adaptive Dual Self-Attention Network for Single Image Denoising

Zuo

Chen²,

Han

et al. 2022

2022 IEEE/CVF Winter Conference on Applications of Computer Vision Workshops (WACVW)

View full text Add to dashboard Cite

show abstract

“…It is better than the comparison algorithm and the algorithm in literature. 15 The article achieves the purpose of improving the positioning accuracy of the RFID multi-tag test system.…”

Section: Localizationmentioning

confidence: 99%

“…Figure 4. After processing the multi-tag image through image matching, we obtained the three-dimensional coordinates of the multi-tag from the horizontal and vertical directions.According to the literature,15 we modeled the relationship between the three-dimensional coordinates on the multi-tag image and the three-dimensional coordinates in the actual space. Finally, we got the position of the multi-tag in the world coordinate system.…”

mentioning

confidence: 99%

A fast RFID image matching algorithm for multi-tag identification and distribution optimization

Zhang

Yu²,

Lin

et al. 2021

Journal of Algorithms & Computational Technology

Self Cite

View full text Add to dashboard Cite

We propose an improved image matching algorithm that combines the minimum feature value algorithm to extract feature points and the direction gradient histogram to calculate the description vector. This algorithm is oriented to RFID multi-tag identification and distribution optimization in the actual scenario, and the traditional SURF algorithm has the problems of low matching accuracy and high complexity in multi-tag matching. This algorithm effectively improves the positioning accuracy of the RFID multi-tag positioning system. The experimental results show that the matching success rate of the improved algorithm in this paper is 87.4%, which is 50% higher than the SURF algorithm. Not only the matching accuracy is greatly improved, but the running speed is also increased by 48%. The algorithm in this paper has high matching accuracy and real-time performance.It provides an effective way for RFID multi-tag real-time fast matching and precise positioning.

show abstract

“…But it must add known blur kernel to multi-tag imaging. In Li et al (2019a), flexible deep convolution neural network denoising model (FDnCNN) of RFID multi-label image was proposed, which removes much noise generated in transmission and acquisition of image. It assumes that all noise is similar to white Gaussian noise (WGN) approximately.…”

Section: Introductionmentioning

confidence: 99%

RFID multi-tag dynamic detection measurement based on conditional generative adversarial nets and YOLOv3

Lin

Liu

et al. 2021

Transactions of the Institute of Measurement and Control

Self Cite

View full text Add to dashboard Cite

The quality of multi-tag imaging greatly affects the effective detection of multi-tag. When multi-tag moves rapidly, the image may have serious dynamic blur, and tags can not be detected efficiently. In previous work, it is generally assumed that blur kernel and noise stationary to improve image quality. However, the dynamic deblurring of Radio Frequency Identification (RFID) multi-tag imaging is an ill-posed inverse problem. In this paper, firstly, blur-sharp multi-tag image pairs are made by superimposing and averaging the adjoin random frames. Then, we propose blind deblurring for dynamic RFID multi-tag imaging based on conditional generative adversarial nets (CGANs), which adds perceptual loss and content loss to generator to make image sharper. Finally, tags are detected by YOLOv3 in real time in end-to-end manner. Experimental results demonstrate that PSNR is at least 0.56dB higher and speed is at least 31.25 % faster than that of the current improved convolution neural networks (CNN). CGANs can remove image blur better, which has great superiority in the field of dynamic multi-tag imaging. In addition, YOLOv3 detects multi-tag quickly, thereby improving the detection accuracy.

show abstract

FDnCNN-based image denoising for multi-labfel localization measurement

Cited by 14 publications

References 32 publications

IDEA-Net: Adaptive Dual Self-Attention Network for Single Image Denoising

IDEA-Net: Adaptive Dual Self-Attention Network for Single Image Denoising

A fast RFID image matching algorithm for multi-tag identification and distribution optimization

RFID multi-tag dynamic detection measurement based on conditional generative adversarial nets and YOLOv3

Contact Info

Product

Resources

About