Sensors

Research

19 pages, 1597 KiB

Open AccessArticle

Effects of JPEG Compression on Vision Transformer Image Classification for Encryption-then-Compression Images

by Genki Hamano, Shoko Imaizumi and Hitoshi Kiya

Sensors 2023, 23(7), 3400; https://doi.org/10.3390/s23073400 - 23 Mar 2023

Cited by 5 | Viewed by 2430

This paper evaluates the effects of JPEG compression on image classification using the Vision Transformer (ViT). In recent years, many studies have been carried out to classify images in the encrypted domain for privacy preservation. Previously, the authors proposed an image classification method [...] Read more.

This paper evaluates the effects of JPEG compression on image classification using the Vision Transformer (ViT). In recent years, many studies have been carried out to classify images in the encrypted domain for privacy preservation. Previously, the authors proposed an image classification method that encrypts both a trained ViT model and test images. Here, an encryption-then-compression system was employed to encrypt the test images, and the ViT model was preliminarily trained by plain images. The classification accuracy in the previous method was exactly equal to that without any encryption for the trained ViT model and test images. However, even though the encrypted test images can be compressible, the practical effects of JPEG, which is a typical lossy compression method, have not been investigated so far. In this paper, we extend our previous method by compressing the encrypted test images with JPEG and verify the classification accuracy for the compressed encrypted-images. Through our experiments, we confirm that the amount of data in the encrypted images can be significantly reduced by JPEG compression, while the classification accuracy of the compressed encrypted-images is highly preserved. For example, when the quality factor is set to 85, this paper shows that the classification accuracy can be maintained at over 98% with a more than 90% reduction in the amount of image data. Additionally, the effectiveness of JPEG compression is demonstrated through comparison with linear quantization. To the best of our knowledge, this is the first study to classify JPEG-compressed encrypted images without sacrificing high accuracy. Through our study, we have come to the conclusion that we can classify compressed encrypted-images without degradation to accuracy. Full article

(This article belongs to the Special Issue Compressed Sensing and Imaging Processing)

► Show Figures

Figure 1

14 pages, 1452 KiB

Open AccessArticle

Compressive Sensing of Medical Images Based on HSV Color Space

by Gandeva Bayu Satrya, I Nyoman Apraz Ramatryana and Soo Young Shin

Sensors 2023, 23(5), 2616; https://doi.org/10.3390/s23052616 - 27 Feb 2023

Cited by 4 | Viewed by 2213

Abstract

Recently, compressive sensing (CS) schemes have been studied as a new compression modality that exploits the sensing matrix in the measurement scheme and the reconstruction scheme to recover the compressed signal. In addition, CS is exploited in medical imaging (MI) to support efficient [...] Read more.

Recently, compressive sensing (CS) schemes have been studied as a new compression modality that exploits the sensing matrix in the measurement scheme and the reconstruction scheme to recover the compressed signal. In addition, CS is exploited in medical imaging (MI) to support efficient sampling, compression, transmission, and storage of a large amount of MI. Although CS of MI has been extensively investigated, the effect of color space in CS of MI has not yet been studied in the literature. To fulfill these requirements, this article proposes a novel CS of MI based on hue-saturation value (HSV), using spread spectrum Fourier sampling (SSFS) and sparsity averaging with reweighted analysis (SARA). An HSV loop that performs SSFS is proposed to obtain a compressed signal. Next, HSV–SARA is proposed to reconstruct MI from the compressed signal. A set of color MIs is investigated, such as colonoscopy, magnetic resonance imaging of the brain and eye, and wireless capsule endoscopy images. Experiments were performed to show the superiority of HSV–SARA over benchmark methods in terms of signal-to-noise ratio (SNR), structural similarity (SSIM) index, and measurement rate (MR). The experiments showed that a color MI, with a resolution of

256 \times 256

pixels, could be compressed by the proposed CS at MR of 0.1, and could be improved in terms of SNR being 15.17% and SSIM being 2.53%. The proposed HSV–SARA can be a solution for color medical image compression and sampling to improve the image acquisition of medical devices. Full article

(This article belongs to the Special Issue Compressed Sensing and Imaging Processing)

► Show Figures

Figure 1

17 pages, 4913 KiB

Open AccessArticle

IEF-CSNET: Information Enhancement and Fusion Network for Compressed Sensing Reconstruction

by Ziqun Zhou, Fengyin Liu and Haibin Shen

Sensors 2023, 23(4), 1886; https://doi.org/10.3390/s23041886 - 8 Feb 2023

Cited by 3 | Viewed by 1526

Abstract

The rapidly growing requirement for data has put forward Compressed Sensing (CS) to realize low-ratio sampling and to reconstruct complete signals. With the intensive development of Deep Neural Network (DNN) methods, performance in image reconstruction from CS measurements is constantly increasing. Currently, many [...] Read more.

The rapidly growing requirement for data has put forward Compressed Sensing (CS) to realize low-ratio sampling and to reconstruct complete signals. With the intensive development of Deep Neural Network (DNN) methods, performance in image reconstruction from CS measurements is constantly increasing. Currently, many network structures pay less attention to the relevance of before- and after-stage results and fail to make full use of relevant information in the compressed domain to achieve interblock information fusion and a great receptive field. Additionally, due to multiple resamplings and several forced compressions of information flow, information loss and network structure redundancy inevitably result. Therefore, an Information Enhancement and Fusion Network for CS reconstruction (IEF-CSNET) is proposed in this work, and a Compressed Information Extension (CIE) module is designed to fuse the compressed information in the compressed domain and greatly expand the receptive field. The Error Comprehensive Consideration Enhancement (ECCE) module enhances the error image by incorporating the previous recovered error so that the interlink among the iterations can be utilized for better recovery. In addition, an Iterative Information Flow Enhancement (IIFE) module is further proposed to complete the progressive recovery with loss-less information transmission during the iteration. In summary, the proposed method achieves the best effect, exhibits high robustness at this stage, with the peak signal-to-noise ratio (PSNR) improved by 0.59 dB on average under all test sets and sampling rates, and presents a greatly improved speed compared with the best algorithm. Full article

(This article belongs to the Special Issue Compressed Sensing and Imaging Processing)

► Show Figures

Figure 1

20 pages, 1111 KiB

Open AccessArticle

Sidelobe Suppression Techniques for Near-Field Multistatic SAR

by George A. J. Price, Chris Moate, Daniel Andre and Peter Yuen

Sensors 2023, 23(2), 732; https://doi.org/10.3390/s23020732 - 9 Jan 2023

Viewed by 1730

Abstract

Multirotor Unmanned Air Systems (UAS) represent a significant improvement in capability for Synthetic Aperture Radar (SAR) imaging when compared to traditional, fixed-wing, platforms. In particular, a swarm of UAS can generate significant measurement diversity through variation of spatial and frequency collections across an [...] Read more.

Multirotor Unmanned Air Systems (UAS) represent a significant improvement in capability for Synthetic Aperture Radar (SAR) imaging when compared to traditional, fixed-wing, platforms. In particular, a swarm of UAS can generate significant measurement diversity through variation of spatial and frequency collections across an array of sensors. In such imaging schemes, the image formation step is challenging due to strong extended sidelobe; however, were this to be effectively managed, a dramatic increase in image quality is theoretically possible. Since 2015, QinetiQ have developed the RIBI system, which uses multiple UAS to perform short-range multistatic collections, and this requires novel near-field processing to mitigate the high sidelobes observed and form actionable imagery. This paper applies a number of algorithms to assess image reconstruction of simulated near-field multistatic SAR with an aim to suppress sidelobes observed in the RIBI system, investigating techniques including traditional SAR processing, regularised linear regression, compressive sensing. In these simulations presented, Elastic net, Orthogonal Matched Pursuit, and Iterative Hard Thresholding all show the ability to suppress sidelobes while preserving accuracy of scatterer RCS. This has also lead to a novel processing approach for reconstructing SAR images based on the observed Elastic net and Iterative Hard Thresholding performance, mitigating weaknesses to generate an improved combined approach. The relative strengths and weaknesses of the algorithms are discussed, as well as their application to more complex real-world imagery. Full article

(This article belongs to the Special Issue Compressed Sensing and Imaging Processing)

► Show Figures

Figure 1

20 pages, 1765 KiB

Open AccessArticle

A Non-Convex Compressed Sensing Model Improving the Energy Efficiency of WSNs for Abnormal Events’ Monitoring

by Yilin Huang, Haiyang Li and Jigen Peng

Sensors 2022, 22(21), 8378; https://doi.org/10.3390/s22218378 - 1 Nov 2022

Cited by 1 | Viewed by 1294

Abstract

The wireless sensor network (WSN), a communication system widely used in the Internet of Things, usually collects physical data in a natural environment and monitors abnormal events. Because of the redundancy of natural data, a compressed-sensing-based model offers energy-efficient data processing to overcome [...] Read more.

The wireless sensor network (WSN), a communication system widely used in the Internet of Things, usually collects physical data in a natural environment and monitors abnormal events. Because of the redundancy of natural data, a compressed-sensing-based model offers energy-efficient data processing to overcome the energy shortages and uneven consumption problems of a WSN. However, the convex relaxation method, which is widely used for a compressed-sensing-based WSN, is not sufficient for reducing data processing energy consumption. In addition, when abnormal events occur, the redundancy of the original data is destroyed, which makes the traditional compressed sensing methods ineffective. In this paper, we use a non-convex fraction function as the surrogate function of the

ℓ_{0}

-norm, which achieves lower energy consumption of the sensor nodes. Moreover, considering abnormal event monitoring in a WSN, we propose a new data construction model and apply an alternate direction iterative thresholding algorithm, which avoids extra measurements, unlike previous algorithms. The results showed that our models and algorithms reduced the WSN’s energy consumption during abnormal events. Full article

(This article belongs to the Special Issue Compressed Sensing and Imaging Processing)

► Show Figures

Figure 1

18 pages, 890 KiB

Open AccessArticle

Construction of Structured Random Measurement Matrices in Semi-Tensor Product Compressed Sensing Based on Combinatorial Designs

by Junying Liang, Haipeng Peng, Lixiang Li and Fenghua Tong

Sensors 2022, 22(21), 8260; https://doi.org/10.3390/s22218260 - 28 Oct 2022

Cited by 2 | Viewed by 1331

Abstract

A random matrix needs large storage space and is difficult to be implemented in hardware, and a deterministic matrix has large reconstruction error. Aiming at these shortcomings, the objective of this paper is to find an effective method to balance these performances. Combining [...] Read more.

A random matrix needs large storage space and is difficult to be implemented in hardware, and a deterministic matrix has large reconstruction error. Aiming at these shortcomings, the objective of this paper is to find an effective method to balance these performances. Combining the advantages of the incidence matrix of combinatorial designs and a random matrix, this paper constructs a structured random matrix by the embedding operation of two seed matrices in which one is the incidence matrix of combinatorial designs, and the other is obtained by Gram–Schmidt orthonormalization of the random matrix. Meanwhile, we provide a new model that applies the structured random matrices to semi-tensor product compressed sensing. Finally, compared with the reconstruction effect of several famous matrices, our matrices are more suitable for the reconstruction of one-dimensional signals and two-dimensional images by experimental methods. Full article

(This article belongs to the Special Issue Compressed Sensing and Imaging Processing)

► Show Figures

Figure 1

16 pages, 4895 KiB

Open AccessArticle

A Novel Reconstruction Algorithm with High Performance for Compressed Ultrafast Imaging

by Qian Shen, Jinshou Tian and Chengquan Pei

Sensors 2022, 22(19), 7372; https://doi.org/10.3390/s22197372 - 28 Sep 2022

Cited by 6 | Viewed by 1534

Abstract

Compressed ultrafast photography (CUP) is a type of two-dimensional (2D) imaging technique to observe ultrafast processes. Intelligence reconstruction methods that influence the imaging quality are an essential part of a CUP system. However, existing reconstruction algorithms mostly rely on image priors and complex [...] Read more.

Compressed ultrafast photography (CUP) is a type of two-dimensional (2D) imaging technique to observe ultrafast processes. Intelligence reconstruction methods that influence the imaging quality are an essential part of a CUP system. However, existing reconstruction algorithms mostly rely on image priors and complex parameter spaces. Therefore, it usually takes a lot of time to obtain acceptable reconstruction results, which limits the practical application of the CUP. In this paper, we proposed a novel reconstruction algorithm named PnP-FFDNet, which can provide a high quality and high efficiency compared to previous methods. First, we built a forward model of the CUP and three sub-optimization problems were obtained using the alternating direction multiplier method (ADMM), and the closed-form solution of the first sub-optimization problem was derived. Secondly, inspired by the PnP-ADMM framework, we used an advanced denoising algorithm based on a neural network named FFDNet to solve the second sub-optimization problem. On the real CUP data, PSNR and SSIM are improved by an average of 3 dB and 0.06, respectively, compared with traditional algorithms. Both on the benchmark dataset and on the real CUP data, the proposed method reduces the running time by an average of about 96% over state-of-the-art algorithms, and show comparable visual results, but in a much shorter running time. Full article

(This article belongs to the Special Issue Compressed Sensing and Imaging Processing)

► Show Figures

Figure 1

13 pages, 45736 KiB

Open AccessArticle

Video Compressive Sensing Reconstruction Using Unfolded LSTM

by Kaiguo Xia, Zhisong Pan and Pengqiang Mao

Sensors 2022, 22(19), 7172; https://doi.org/10.3390/s22197172 - 21 Sep 2022

Cited by 6 | Viewed by 1549

Abstract

Video compression sensing can use a few measurements to obtain the original video by reconstruction algorithms. There is a natural correlation between video frames, and how to exploit this feature becomes the key to improving the reconstruction quality. More and more deep learning-based [...] Read more.

Video compression sensing can use a few measurements to obtain the original video by reconstruction algorithms. There is a natural correlation between video frames, and how to exploit this feature becomes the key to improving the reconstruction quality. More and more deep learning-based video compression sensing (VCS) methods are proposed. Some methods overlook interframe information, so they fail to achieve satisfactory reconstruction quality. Some use complex network structures to exploit the interframe information, but it increases the parameters and makes the training process more complicated. To overcome the limitations of existing VCS methods, we propose an efficient end-to-end VCS network, which integrates the measurement and reconstruction into one whole framework. In the measurement part, we train a measurement matrix rather than a pre-prepared random matrix, which fits the video reconstruction task better. An unfolded LSTM network is utilized in the reconstruction part, deeply fusing the intra- and interframe spatial–temporal information. The proposed method has higher reconstruction accuracy than existing video compression sensing networks and even performs well at measurement ratios as low as 0.01. Full article

(This article belongs to the Special Issue Compressed Sensing and Imaging Processing)

► Show Figures

Figure 1

12 pages, 8991 KiB

Open AccessArticle

WER-Net: A New Lightweight Wide-Spectrum Encoding and Reconstruction Neural Network Applied to Computational Spectrum

by Xinran Ding, Lin Yang, Mingyang Yi, Zhiteng Zhang, Zhen Liu and Huaiyuan Liu

Sensors 2022, 22(16), 6089; https://doi.org/10.3390/s22166089 - 15 Aug 2022

Cited by 3 | Viewed by 1505

Abstract

The computational spectrometer has significant potential for portable in situ applications. Encoding and reconstruction are the most critical technical procedures. In encoding, the random mass production and selection method lacks quantitative designs which leads to low encoding efficiency. In reconstruction, traditional spectrum reconstruction [...] Read more.

The computational spectrometer has significant potential for portable in situ applications. Encoding and reconstruction are the most critical technical procedures. In encoding, the random mass production and selection method lacks quantitative designs which leads to low encoding efficiency. In reconstruction, traditional spectrum reconstruction algorithms such as matching tracking and gradient descent demonstrate disadvantages like limited accuracy and efficiency. In this paper, we propose a new lightweight convolutional neural network called the wide-spectrum encoding and reconstruction neural network (WER-Net), which includes optical filters, quantitative spectral transmittance encoding, and fast spectral reconstruction of the encoded spectral information. The spectral transmittance curve obtained by WER-net can be fabricated through the inverse design network. The spectrometer developed based on WER-net experimentally demonstrates that it can achieve a 2-nm high resolution. In addition, the spectral transmittance encoding curve trained by WER-Net has also achieved good performance in other spectral reconstruction algorithms. Full article

(This article belongs to the Special Issue Compressed Sensing and Imaging Processing)

► Show Figures

Figure 1

12 pages, 2023 KiB

Open AccessCommunication

Adaptive Modular Convolutional Neural Network for Image Recognition

by Wenbo Wu and Yun Pan

Sensors 2022, 22(15), 5488; https://doi.org/10.3390/s22155488 - 22 Jul 2022

Cited by 11 | Viewed by 2224

Abstract

Image recognition has long been one of the research hotspots in computer vision tasks. The development of deep learning is rapid in recent years, and convolutional neural networks usually need to be designed with fixed resources. If sufficient resources are available, the model [...] Read more.

Image recognition has long been one of the research hotspots in computer vision tasks. The development of deep learning is rapid in recent years, and convolutional neural networks usually need to be designed with fixed resources. If sufficient resources are available, the model can be scaled up to achieve higher accuracy, for example, VggNet, ResNet, GoogLeNet, etc. Although the accuracy of large-scale models has been improved, the following problems will occur with the expansion of model scale: (1) There may be over-fitting; (2) increasing model parameters; (3) slow model convergence. This paper proposes a design method for a modular convolutional neural network model which solves the problem of over-fitting and large model parameters by connecting multiple modules in parallel. Moreover, each module contains several submodules (three submodules in this paper) and fuses the features extracted from the submodules. The model convergence can be accelerated by using the fused features (the fused features contain more image information). In this study, we add a gate unit based on the attention mechanism to the model, which aims to optimize the structure of the model (select the optimal number of modules), allowing the model to select an optimum network structure by learning and dynamically reducing FLOPs (floating-point operations per second) of the model. Compared to VggNet, ResNet, and GoogLeNet, the structure of the model proposed in this paper is simple and the parameters are small. The proposed model achieves good results in the Kaggle datasets Cats-vs.-Dogs (99.3%), 10-Monkey Species (99.26%), and Birds-400 (99.13%). Full article

(This article belongs to the Special Issue Compressed Sensing and Imaging Processing)

► Show Figures

Figure 1

24 pages, 6291 KiB

Open AccessArticle

Low-Complexity Adaptive Sampling of Block Compressed Sensing Based on Distortion Minimization

by Qunlin Chen, Derong Chen and Jiulu Gong

Sensors 2022, 22(13), 4806; https://doi.org/10.3390/s22134806 - 25 Jun 2022

Cited by 3 | Viewed by 1443

Abstract

Block compressed sensing (BCS) is suitable for image sampling and compression in resource-constrained applications. Adaptive sampling methods can effectively improve the rate-distortion performance of BCS. However, adaptive sampling methods bring high computational complexity to the encoder, which loses the superiority of BCS. In [...] Read more.

Block compressed sensing (BCS) is suitable for image sampling and compression in resource-constrained applications. Adaptive sampling methods can effectively improve the rate-distortion performance of BCS. However, adaptive sampling methods bring high computational complexity to the encoder, which loses the superiority of BCS. In this paper, we focus on improving the adaptive sampling performance at the cost of low computational complexity. Firstly, we analyze the additional computational complexity of the existing adaptive sampling methods for BCS. Secondly, the adaptive sampling problem of BCS is modeled as a distortion minimization problem. We present three distortion models to reveal the relationship between block sampling rate and block distortion and use a simple neural network to predict the model parameters from several measurements. Finally, a fast estimation method is proposed to allocate block sampling rates based on distortion minimization. The results demonstrate that the proposed estimation method of block sampling rates is effective. Two of the three proposed distortion models can make the proposed estimation method have better performance than the existing adaptive sampling methods of BCS. Compared with the calculation of BCS at the sampling rate of 0.1, the additional calculation of the proposed adaptive sampling method is less than 1.9%. Full article

(This article belongs to the Special Issue Compressed Sensing and Imaging Processing)

► Show Figures

Figure 1

Journal Menu

Journal Browser

Compressed Sensing and Imaging Processing

Share This Special Issue

Special Issue Editor

Special Issue Information

Related Special Issue

Published Papers (11 papers)

Research

Further Information

Guidelines

MDPI Initiatives

Follow MDPI