Image Data Transmission Research Articles

Deep joint source-channel coding empowered two-way relay networks for wireless image transmission

Compared with the traditional uni-directional relaying, two-way relay networks provide important enhancements and optimizations to modern communication systems. However, with the increasing requirements of artificial intelligence applications for image data transmission, relay-assisted communication technologies are reaching the theoretical limit in terms of bandwidth, which hinders the further development of AI applications. To address this issue, we propose a deep joint source-channel coding empowered two-way relay network (DeepJSCC-TWRN) to help image transmission. Specifically, in the DeepJSCC-TWRN, a DeepJSCC is employed to improve image transmission quality of the TWRN from the perspective of visual semantic information, and each source can achieve optimal performance by being trained in a uniform deep learning framework. For measuring the performance of the proposed DeepJSCC-TWRN, we employ the peak signal-to-noise ratio (PSNR) and the structural similarity index measure (SSIM) as performance metrics. Simulation results show that DeepJSCC-TWRN outperforms the baseline method, demonstrating the ability to preserve visual semantic information.

Specular highlight removal by federated generative adversarial network with attention mechanism

Specular highlight removal ensures the acquisition of high-quality images, which finds its important applications in stereo matching, text recognition and image segmentation. In order to prevent the leakage of images containing personal information, such as identification card (ID) photos, clients often train specular highlight removal models using local data resulting in a lack of precision and generalization of the trained model. To address this challenge, this paper introduces a new method to remove highlight in images using federated learning (FL) and attention generative adversarial network (AttGAN). Specifically, the former builds a global model in the central server and updates the global model by aggregating model parameters of clients. This process does not involve the transmission of image data, which enhances the privacy of clients; the later combining attention mechanisms and generative adversarial network aims to improve the quality of highlight removal by focusing on key image regions, resulting in more realistic and visually pleasing results. The proposed FL-AttGAN method is numerically evaluated, using SD1, SD2 amd RD datasets. The results show that the proposed FL-AttGAN outperforms existent methods.

Precision encoder grating mounting: a near-sensor computing approach.

The rotary motor plays a pivotal role in various motion execution mechanisms. However, an inherent issue arises during the initial installation of the encoder grating, namely, eccentricity between the centers of the encoder grating and motor shaft. This eccentricity substantially affects the accuracy of motor angle measurements. To address this challenge, we proposed a precision encoder grating mounting system that automates the encoder grating mounting process. The proposed system mainly comprises a near-sensor detector and a push rod. With the use of a near-sensor approach, the detector captures rotating encoder grating images, and the eccentricity is computed in real-time. This approach substantially reduces the time delays in image data transmission, thereby enhancing the speed and accuracy of eccentricity calculation. The major contribution of this article is a method for real-time eccentricity calculation that leverages an edge processor within the detector and an edge-vision baseline detection algorithm. This method enables real-time determination of the eccentricity and eccentricity angle of the encoder grating. Leveraging the obtained eccentricity and eccentricity angle data, the position of the encoder grating can be automatically adjusted by the push rod. In the experimental results, the detector can obtain the eccentricity and eccentricity angle of the encoder grating within 2.8 s. The system efficiently and precisely completes a encoder grating mounting task in average 25.1 s, and the average eccentricity after encoder grating mounting is 3.8 µm.

Pathological correlation between eosinophils and thyroid nodules based on medical image testing

Thyroid nodule is a common thyroid disease, but the study of its pathology and pathogenesis is still limited. As a non-invasive diagnostic method, medical image examination is of great value to study the pathological correlation of thyroid nodules. The purpose of this study was to investigate the expression of eosinophils in medical image examination and the pathological correlation between eosinophils and thyroid nodules. The study analyzed the pathological reports of a group of patients with thyroid nodules examined by medical images and performed corresponding imaging scans or examinations. The imaging data is processed, including image reconstruction, data transmission and other steps, to generate images that can be diagnosed by doctors. Thyroid function and parameters of leukocyte were collected and compared.The serum levels of TT4 and fT4 were observed lower in G2 group, while thyroid stimulating hormone (TSH) was higher compared to G1 group before surgery. Compared to G2 group, eosinophils count and percentage were lower in G1group (p < 0.05) post-surgery and lower ratio of eosinophils count with lymphocyte count (ELR) were observed in G1 group patients (p < 0.05).Elevated TSH is closely related to malignant TN per surgery, while lower ELR suggesting that TN removed thoroughly. Relevant cut-off values required further study to guide the diagnosis, treatment and follow-up of TN.

Radiation Anomaly Detection of Sub-Band Optical Remote Sensing Images Based on Multiscale Deep Dynamic Fusion and Adaptive Optimization

Radiation anomalies in optical remote sensing images frequently occur due to electronic issues within the image sensor or data transmission errors. These radiation anomalies can be categorized into several types, including CCD, StripeNoise, RandomCode1, RandomCode2, ImageMissing, and Tap. To ensure the retention of image data with minimal radiation issues as much as possible, this paper adopts a self-made radiation dataset and proposes a FlexVisionNet-YOLO network to detect radiation anomalies more accurately. Firstly, RepViT is used as the backbone network with a vision transformer architecture to better capture global and local features. Its multiscale feature fusion mechanism efficiently handles targets of different sizes and shapes, enhancing the detection ability for radiation anomalies. Secondly, a feature depth fusion network is proposed in the Feature Fusion part, which significantly improves the flexibility and accuracy of feature fusion and thus enhances the detection and classification performance of complex remote sensing images. Finally, Inner-CIoU is used in the Head part for edge regression, which significantly improves the localization accuracy by finely adjusting the target edges; Slide-Loss is used for classification loss, which enhances the classification robustness by dynamically adjusting the category probabilities and markedly improves the classification accuracy, especially in the sample imbalance dataset. Experimental results show that, compared to YOLOv8, the proposed FlexVisionNet-YOLO method improves precision, recall, mAP0.5, and mAP0.5:0.9 by 3.5%, 7.1%, 4.4%, and 13.6%, respectively. Its effectiveness in detecting radiation anomalies surpasses that of other models.

Solving satellite image data downlink scheduling problem with family attribute via a bi-stage differential evolutionary algorithm

Due to the asynchronous development of observation and transition capabilities, the original image data (OID) obtained by a one-time observation cannot be completely transmitted in one transmission cycle between the EOS and GS, which is named a visible time window (VTW). First, it is needed to segment the OID into several segmented image datasets (SIDs), which are then transmitted in several VTWs. Furthermore, this can cause the satellite image data downlink scheduling problem (SIDSP). This study defines an innovative SIDSP as a satellite image data downlink scheduling problem with the family attribute (SIDSPFA), where massive OID are segmented using a fast segmentation operator, and all SIDs and other no-segmented OID are transmitted in the second step. In addition, two optimization objectives, namely the failure rate (FR) and Segmentation Time (ST) of the image data transmission, are defined to formalize the SIDSPFA as a bi-objective discrete optimization model. Further, a bi-stage differential evolutionary algorithm (DE-CD) is developed. The results of the extensive simulation tests verify the efficiency of the models, strategies, algorithms, and operators used in this study.

Enhancing information transmission in FSO-OAM systems through multiscale interpretable neural networks under turbulent conditions

The paper proposes a solution to improve the information transmission efficiency of FSO-OAM systems under turbulent conditions by combining a multiscale interpretable neural network model, 4RK-MSNN. We use a multiscale structure to design the overall architecture of the neural network, which enables the comprehensive analysis of information in different dimensions. Based on the fourth-order Runge-Kutta correlation theory, a core network module, 4RK, is constructed, which can be explained in terms of dynamical systems. The 4RK-MSNN model, which couples the multiscale structure and the 4RK module, has a lower number of parameters, allowing for layered feature extraction in an interpretable framework. This facilitates low-cost, rapid sharing and transmission of feature information at different scales. The proposed solution is validated by transmitting image data under different turbulence intensities and transmission distances. The results indicate the feasibility of the proposed information transfer system. After adding redundant training data, the 4RK-MSNN model significantly improves the quality of the transmitted data and maintains satisfactory results even under strong turbulence and long distances.

Tracking the transmission matrix of a moving fiber with the transmitted data.

During data transmission, the dynamic change of a scattering medium will make the measured transmission matrix (TM) invalid, so it is necessary to repeatedly measure the TM to achieve a long-time data transmission, which requires stopping the data transmission process frequently to measure the TM and leads to a reduction in the communication capacity. To solve this problem, we propose a TM tracking method during data transmission. In the case of more than three discrete levels of phase modulation, this method can realize the calibration of the TM with the intensity pictures captured by the camera and the recovered data, so it does not require stopping the data transmission process to measure the TM and thus avoids the loss of communication capacity. We have proved the feasibility of this method through simulations and experiments and realized the continuous transmission of random data and image data through a moving fiber with high accuracy.

A new Sylvester-type quaternion matrix equation model for color image data transmission

A new Sylvester-type quaternion matrix equation model for color image data transmission

Image encryption scheme based on improved four-dimensional chaotic system and evolutionary operators

To enhance the security of image data transmission, and address the weaknesses of existing image encryption schemes based on chaotic systems, particularly concerning resistance to differential attacks and the unstable performance of chaotic systems, this article introduces an improved four-dimensional chaotic system and integrates evolutionary operators to propose an image encryption scheme. Firstly, a method for generating pseudo-random sequences associated with the plaintext is designed. The change rate of the ciphertext pixel value exceeds 0.9967 after a slight modification of the plaintext pixel value, significantly improving the plaintext sensitivity and the scheme's ability to resist selected plaintext attacks. Secondly, an individual rearrangement operation is introduced to achieve bit-level scrambling, and pixel-level scrambling is achieved by selection strategy. Subsequently, crossover and mutation operations are incorporated into image encryption. To reflect the randomness of the pairing, we adopt the pseudo-random sequence generated by the chaotic system to control the crossover and mutation operators, and a diffusion operation is performed on selected pixel pairs. Finally, ciphertext feedback is applied. Experimental results and performance analysis demonstrate that the proposed scheme not only enhances the security of encrypted images but also effectively resists noise and cropping attacks. This method effectively meets the high-security requirements of images in network transmission and provides new ideas for further research in the field of image encryption.

High-Resolution Image Processing and Spatiotemporal Data Transmission System Based on GPU Acceleration

With the development of information technology and the increasing demand for data processing, the serial mode of the central processing unit (CPU) is difficult to efficiently transmit large-scale spatiotemporal data, and the processing effect for high-resolution images is not good. This paper designed a high-resolution image processing and spatiotemporal data transmission system based on graphics processing unit (GPU) acceleration to improve the processing efficiency of large-scale spatiotemporal data. In this paper, traffic spatiotemporal data was taken as an example for analysis. Large-scale traffic image data was collected by road monitoring equipment, and image compression was performed on the collected image. Fourier transform was used to eliminate image data redundancy, and GPU-accelerated parallel processing was used to achieve fast image defogging and data transmission. This paper selected 2TB of traffic spatiotemporal data with image resolutions of 540P, 720P, 1080P, 1440P, and 2160P. GPU acceleration was performed using the Compute Unified Device Architecture (CUDA). In images with a resolution of 2160P, the processing time for CPU and GPU acceleration was 2900ms and 28ms, respectively, with an acceleration ratio of 103.6. A high-resolution image processing and spatiotemporal data transmission system based on GPU acceleration can improve the efficiency of traffic spatiotemporal data processing and have excellent concurrent processing capabilities.

Optimization of Image Transmission in Cooperative Semantic Communication Networks

In this paper, a semantic communication framework for image data transmission is developed. In the investigated framework, a set of servers cooperatively transmit image data to a set of users utilizing semantic communication techniques, which enable servers to transmit only the semantic information that accurately captures the meaning of images. To evaluate the performance of studied semantic communication system, a multimodal metric called image-to-graph semantic similarity (ISS) is proposed to measure the correlation between the extracted semantic information and the original image. To meet the ISS requirement of each user, each server must jointly determine the semantic information to be transmitted and the resource blocks (RBs) used for semantic information transmission. Due to the co-channel interference among users associated with different servers, each server must cooperate with other servers to find a globally optimal semantic oriented RB allocation. We formulate this problem as an optimization problem whose goal is to minimize the sum of the average transmission latency of each server while reaching the ISS requirement. To solve this problem, we propose a value decomposition based entropy-maximized multi-agent reinforcement learning (RL) algorithm. The proposed algorithm enables each server to coordinate with other servers in training stage and execute RB allocation in a distributed manner to approach to a globally optimal performance with less training iterations. Compared to traditional multi-agent RL algorithms, the proposed RL framework improves the exploration of valuable action of servers and the probability of finding a globally optimal RB allocation policy based on local observation of wireless and semantic communication environments. Simulation results show that the proposed algorithm can reduce the transmission delay by up to 16.1% and improve the convergence speed by up to 100% compared to the traditional multi-agent RL algorithms.

White rice stem borer pest detection system using image-based convolution neural network

Preventing agricultural resource loss caused by pests remains a crucial issue. While technological advancements are being achieved, the current agricultural management methods and equipment have yet to meet the required level for precise pest control, a huge portion of the pest population analysis process is still conducted manually. As a solution to this issue, the development of a White Rice Stem Borer pest detection system has been conducted by applying Convolutional Neural Network (CNN) technology to calculate the pest population count at the research location. This system has been specifically designed to detect the White Rice Stem Borer using available traps. The method involves training data from a direct dataset obtained from the field, categorized into two positive and negative classes of the White Stem Borer pests. Six models have been trained from this dataset, utilizing two different architectures. Out of the six trained models, four showed potential overfitting, one exhibited underfitting, and one model demonstrated optimal results. The highest accuracy in image detection achieved by the most optimal CNN model was 97.35%, with a training accuracy of 98.54%. This best-performing model utilized an architecture with three Convolution layers, 50 Epochs, and an automatic data split with an 80:20 training-validation data ratio. From the research findings, it is concluded that this study can assist in automatically analyzing the quantity of White Stem Borer pests in a specific area without directly counting the number of pests from existing traps. However, the study still encounters a limitation—the detection process still requires substantial server resources and cannot be directly processed on the Raspberry PI device installed in the trap. Consequently, the detection relies on transmitting image data from the field device to the server before the detection process can occur.

Revolutionizing Crop Monitoring: AI-Driven Machine Vision for Real-time Adaptive Harvesting and Loss Mitigation

The research on implementing management strategies for crops that utilize the ESP32-CAM microcontroller to mitigate losses during the harvesting process. The capabilities of the ESP32-CAM in image processing and data transmission are utilized to construct a live monitoring system for agricultural fields. Using image analysis and machine learning algorithms, the system can identify the maturity of crops, the presence of pests, and the environmental conditions. It then provides practical advice to help make timely decisions about when to harvest. In addition, the ESP32-CAM serves as a tool for remote surveillance and management, enhancing the efficiency of farmers' activities and enabling the mitigation of losses during harvesting. The research highlights the integration of IoT technology with crop management techniques in contemporary precision agriculture, resulting in improved efficiency and sustainability of agricultural production.

An efficient and lightweight image encryption technique using Lorenz chaotic system

In the past few years, to store and transmit image data securely, numerous research initiatives on image encoding have been conducted. The primary objective of the image encryption technique is to safeguard the image by sabotaging the pixel pattern. Researchers suggested a safe, portable, and simple to use image encryption technique in this work. The encryption of the image is done using a bit-wise XOR operation, where the bit-wise operation is applied on each pixel of the plain image with a pseudo-random number that is created by the Lorenz chaotic system, to prevent unwanted access to confidential image data. The results of the experiments demonstrate that the suggested technique offers effective image encryption and decryption. The key stream of the encrypted image is made up of pseudo-random digits generated by the Lorenz Chaotic System. Several experimental tests have been performed, including histogram, correlation, information entropy, and differential analysis. The experimental findings reveal that the suggested approach performs image encryption and decryption efficiently.

Hop count, ETX and energy selection-based objective function for image data transmission over 6LoWPAN in Internet of Things

Hop count, ETX and energy selection-based objective function for image data transmission over 6LoWPAN in Internet of Things

Hop count, ETX and energy selection-based objective function for image data transmission over 6LoWPAN in Internet of Things

Hop count, ETX and energy selection-based objective function for image data transmission over 6LoWPAN in Internet of Things

A reversible and lossless secret image sharing scheme with authentication for color images

With the increasing concern for public privacy protection, Secret Image Sharing (SIS) has garnered significant attention among researchers. The goal of SIS is to split a secret image into multiple shadows, rendering it unrecoverable unless a specific threshold of shares is obtained. However, these shadows typically appear as meaningless noise-like patterns, making them susceptible to attacks and raising suspicion when transmitted over traditional internet channels. Moreover, traditional SIS schemes often restrict image formats to binary, limiting their applicability. To address these limitations, we propose a reversible and lossless secret image sharing scheme with authentication specifically designed for color images. Our scheme generates noise-like shadows by combining original secret images and authentication images, which are then embedded into color images using an enhanced steganography method. During the recovery process, participants can extract the shares losslessly and complete the authentication phase without the need for a trusted third party. The authentication results can be easily verified by visual inspection. Ultimately, the qualified shadows are utilized to reconstruct the secret image, ensuring secure transmission of confidential image data. The efficacy and security of our proposed method are demonstrated via theoretical analysis and experimental results.

Assessing efficacy of cellular transmission technology in camera trapping for wildlife research

AbstractCamera traps are an important noninvasive tool used by scientists to monitor wildlife efficiently and at reduced costs. New camera trap features improve performance and encourage increased use by researchers and the public. Cellular transmission of image data, which provides users the ability to digitally receive images instead of retrieving or downloading images in the field is a useful new feature. Cellular data transmission has 2 key benefits for wildlife research in that it reduces travel time required for downloading image data and the uncertainty involving storage capacity of SD cards and battery life, and cellular transmission allows for near real‐time analysis of images, which could redistribute the time usually devoted to processing a large data set when the memory card is retrieved. Despite potential benefits, cellular transmission technology in camera traps is still new and questions remain about its reliability. Our objective was to determine the efficacy of cellular transmission technology in wildlife research by designing a camera trap study as part of a senior‐level class (Wildlife Management Applications and Planning; NREM 4522) project at Oklahoma State University. We used ArcGIS to generate a stratified random sample of trap locations, deploying five cellular transmission camera traps in open grassland and five in closed canopy forest areas from 5 September to 5 October 2021. We monitored the number of transmitted images each day online, and after camera trap retrieval, we compared the number of transmitted images to those stored on the memory card to determine transmission efficiency. Our data indicated the majority of the images taken each day were transmitted successfully; however, transmission efficiency (i.e., number transmitted divided by total number taken by the camera trap) tended to be lower in forested areas (47%) compared to open grassland (86%). Though cellular transmission technology shows promise, the combination of cellular signal, landscape features, and transmitted data quality may limit the effectiveness of cellular transmission technology for near real‐time data analysis. Based on our results, we recommend that researchers consider advantages and disadvantages of cellular transmission when designing studies and note that researchers may need to adopt an adaptive approach or conduct pilot testing that includes quantifying the transmission functionality.

CIS-based system for acquiring and transmitting image data

The design utilizes a network of CIS lines (DL520-09UHM-S) to design an image data acquisition and transmission system. Using the STM32F1 chip as the control core, the hardware circuit of STM32F1 with DL520-09UHM-S, A/D converter chip, FIFO cache chip and CAN transceiver chip is designed, and the parameters of each device are controlled to complete the acquisition and transmission of image data. The PC processes the data through software to derive the acquired image data.