Pixel Points Research Articles

CONNECTING COMMUNITIES: ENHANCING BROADBAND ACCESS IN RURAL TANZANIA THROUGH SMALL CELL DEPLOYMENT

Aim: The aim of this study is to explores potential performance enhancement to a network infrastructure comprising with Macro-cells only by introducing and deploying cost-effective and adaptable small cell backhaul connections. For the marginalized rural areas and terrain varied areas, the aim is to explore how small cell deployment can bridge this gap and improve connectivity for underserved communities. Methodology: A framework is designed by leveraging LTE-Advanced features and integrating Macro cells and small cells in WinProp tool for wireless propagation and radio network planning. This proposed methodology is analysed through system simulations of a carefully selected practical deployment scenario in Southern Chunya, Tanzania. The path loss data of each pixel point in a study area is collected through this simulation study. The performance metrics of cellular signal transmission such as throughput, signal to interference noise ratio and spectral efficiency are examined through MATLAB tool. Results: The results illustrate that the deployment of small cells with LTE-Advanced features can enhance network coverage and overall quality, particularly in rural regions. The performance metrics of cellular signal transmission such as throughput, signal to interference noise ratio and spectral efficiency applied under this study have all shown improvement for a framework comprises with both macro cells and small cells as compared to a framework with Macro-cell alone. More over the results shows there is better results of throughput at 800MHz frequency band as compared to 2000MHz by little margin. The results were verified by conducting comparative assessment of similar studies. Conclusion The study has shown that small cells are potential for improving last mile connectivity and backhauling. The findings in this study will significantly guide future research and initiatives to bridge the digital divide and promote equitable access to broadband services in Tanzania and similar contexts worldwide.

An adaptive image compression algorithm based on joint clustering algorithm and deep learning

AbstractIn recent years, deep artificial neural networks have attracted much attention and have been applied in various fields because they surpass the parameter fitting effect of traditional methods under the condition of data convergence. On the other hand, limited transmission bandwidth and storage capacity make image compression necessary in communication. Here, a compression algorithm that combines the K‐means clustering algorithm with the neural network algorithm is proposed. First, the pixel points of the image are clustered by K‐means algorithm in order to reduce the amount of data input to the neural network algorithm. Secondly, neural network is used to extract image features which realizes further compression. The experiment results show that the peak signal‐to‐noise ratio (PSNR) is 33.48 dB at most with compression ratio at 32:1. The ablation experiment shows that the run time speeds up 9.5% compared to the algorithm without K‐means clustering. Comprehensive comparison experiment shows that the average PSNR is 30.09 dB, which is larger than other baseline approaches. The proposed algorithm is an efficient solution for image compression.

LFSR state sequence image encryption method based on VHDL language

In the modern society of digitalization, integration, intelligence, and networking, while people enjoy the convenience of information technology to their production and lives, information security in the network, as the cornerstone of information communication, becomes more and more important. The research is to establish a new image encryption (IE) method based on LFSR state sequence (SS)s in VHDL language, and the stream cipher of LFSR is studied in detail to induce the idea of LFSR SSs based on VHDL language. The correlation coefficient (CC) of the original image (OI) and encrypted image (EI) pixel points (PP) are analyzed from horizontal direction, vertical direction and diagonal direction, and the results show that the CC of adjacent PP of the OI is large, which approaches 1. However, using the encryption algorithm proposed in this paper, the correlation coefficient of the PP of the EI is -0.0282 in the diagonal direction, and the highest correlation coefficient in the horizontal direction is only 0.0122, which indicates that the adjacent PP of the EI are almost uncorrelated with each other. Therefore, the encryption method can well resist statistical attacks, which illustrates the effectiveness, security, and reliability of this new IE method.

Augmented Grad-CAM++: Super-Resolution Saliency Maps for Visual Interpretation of Deep Neural Network

In recent years, deep neural networks have shown superior performance in various fields, but interpretability has always been the Achilles’ heel of deep neural networks. The existing visual interpretation methods for deep neural networks still suffer from inaccurate and insufficient target localization and low-resolution saliency maps. To address the above issues, this paper presents a saliency map generation method based on image geometry augmentation and super-resolution called augmented high-order gradient weighting class activation mapping (augmented grad-CAM++). Unlike previous approaches that rely on a single input image to generate saliency maps, this method first introduces the image geometry augmentation technique to create a set of augmented images for the input image and generate activation mappings separately. Secondly, the augmented activation mappings are combined to form the final saliency map. Finally, a super-resolution technique is introduced to add pixel points to reconstruct the saliency map pixels to improve the resolution of the saliency map. The proposed method is applied to analyze standard image data and industrial surface defect images. The results indicate that, in experiments conducted on standard image data, the proposed method achieved a 3.1% improvement in the accuracy of capturing target objects compared to traditional methods. Furthermore, the resolution of saliency maps was three times higher than that of traditional methods. In the application of industrial surface defect detection, the proposed method demonstrated an 11.6% enhancement in the accuracy of capturing target objects, concurrently reducing the false positive rate. The presented approach enables more accurate and comprehensive capture of target objects with higher resolution, thereby enhancing the visual interpretability of deep neural networks. This improvement contributes to the greater interpretability of deep learning models in industrial applications, offering substantial performance gains for the practical deployment of deep learning networks in the industrial domain.

Universal method using a pre-deformed reference subset to eliminate the interpolation bias in digital image correlation

The high measurement accuracy of the digital image correlation (DIC) method is derived from the sub-pixel registration algorithm, which interpolates the intensities at the sub-pixel position in the image. The displacement error caused by the interpolation is a systematic bias in the DIC method, known as the sinusoidal bias in the sub-pixel translation experiment. Although the interpolation bias has been well researched, there is a lack of a universal method to eliminate interpolation bias. In this work, we propose a universal method to eliminate the interpolation bias using a pre-deformed reference subset; pixel points in the pre-deformed subset are deviated from the integer-pixel location. The purpose of the adjustment is to set the deformed pixel points at a specific position, so that the interpolation bias of all deformed pixel points cancels each other out, close to zero. The adjustment of the pre-deformed reference subset is related with the subset size and subset deformation. Numerical experiments including DIC challenge data and a real uniaxial tensile test were conducted to verify the effectiveness and universality of the proposed method, contributing to improved measurement accuracy. Considering the effect of pixel point location on the interpolation bias, this work proposes a universal method to eliminate the interpolation bias and provides a perspective to study DIC errors.

Camera calibration method through multivariate quadratic regression for depth estimation on a stereo vision system

Machine vision systems have demonstrated practical applications in various fields where the perception of the environment is essential. Stereo vision systems are one of the most used methods to perform three-dimensional mapping. These systems have several considerations in order to fulfill this task; One of them is camera calibration. This is why this paper proposes a novel camera calibration method for improving depth estimation accuracy in stereo vision systems. The method addresses the lens distortion issue by adjusting the surface point's pixel coordinates before the triangulation process. The new pixel coordinates are computed using calibration coefficients, which are calculated through multivariate quadratic regression. Also, the proposed method corrects the relative orientation of the cameras and computes a compensation angle. The proposed method is remarkably robust to varying illumination levels because it adjusts pixel positions independently of image content. In comparison to traditional methods, this method requires fewer steps to implement, as it only requires one image per camera. The proposed method could have significant implications for fields such as autonomous navigation and robotics, where depth estimation is a crucial component. Overall, this paper presents a valuable contribution to the field of computer vision by offering a new, efficient and effective approach to camera calibration for stereo vision systems. The performed experiments demonstrated that the proposed method leads to improved depth estimation in the stereo vision system with an improvement of 34.15% on the MAE and 48.38% on the STD when compared to one of the most commonly used methods.

Teaching Chinese Painting Colour Based on Intelligent Image Processing Technology

Abstract This paper constructs a Chinese painting color teaching platform based on intelligent image color processing technology. Firstly, the region segmentation method is used to reasonably segment the pixel points according to the similar values of color and texture parameters of the image. Then, the image’s primary color grayscale parameters are fuzzy, detected, equalized, and fused by color difference values. Finally, the gradient optimization algorithm is combined to identify and control the image color parameters. The results show that the highest value of peak signal-to-noise ratio can reach 73.4db for the images processed by the method of this paper, and the mean value of STRESS is kept between 0-10%. The intelligent image color processing technique has been proven to enhance the aesthetics and innovation of Chinese painting colors in students.

Analysis of the expression and application of visual communication elements in art design based on big data

Abstract Based on big data analysis, this paper clarifies the basic elements and other elements in visual communication through the expression of visual communication elements in art design and constructs the formation mechanism of visual art design and the characteristics of information cognition. Based on the image processing algorithm for image brightness equalization processing, the image fusion and wavelet noise reduction processing are combined with the pixel point quantization tracking method, and specific cases demonstrate the optimization effect and practicality. The results show that in the example solution, the visual communication index of each module is Z = 4289, while the optimized visual communication index Z&prime = 4834 is obtained after processing by the image-based processing algorithm, which improves the communication effect and quality of the art design. This paper proposes a new idea and method that provides inspiration and reference for art design practice and research.

Binocular Ranging Design Based on FPGA

In daily life, distance measurement is everywhere. In order to accurately measure the distance information of the target object, a system design for binocular ranging based on FPGA is proposed. The internal and external parameters of the binocular camera are determined through the Zhang Zhengyou calibration method (two-dimensional target), and then the basic model of the binocular camera is derived. The data model of the binocular camera is derived based on the binocular camera model. The calibrated binocular camera parameters are used to perform binocular correction and distortion correction on the image data collected by the binocular camera. The corrected left and right eye views are matched with pixel points through a stereo matching algorithm to form a disparity map. The principle of triangle similarity is used to estimate the target distance and obtain the coordinate system of the target. Finally, the data is measured by deploying it on the FPGA development board.

Stretchable, Programmable and Magnet-Insensitive Protonic Display Based on Integrated Ionic Circuit.

As a new approach to "More than Moore", integrated ionic circuits serve as a possible alternative to traditional electronic circuits, yet the integrated ionic circuit composed of functional ionic elements and ionic connections is still challenging. Herein, a stretchable and transparent ionic display module of the integrated ionic circuit has been successfully prepared and demonstrated by pixelating a proton-responsive hydrogel. It is programmed to excite the hydrogel color change by a Faraday process occurring at the electrode at the specific pixel points, which enables the display of digital information and even color information. Importantly, the display module exhibits stable performance under strong magnetic field conditions (1.7T). The transparent and stretchable nature of such ionic modules also allows them to be utilized in a broad range of scenarios, which paves the way for integrated ionic circuits.

Quantitative analysis method for infrared characterization of coal damage under load based on image enhancement and pixel extraction

To explore the quantification method of infrared thermal image characteristics in the coal loaded-damage process and better reflect the infrared precursor information and damage evolution of coal, in this paper, a new index based on the color extraction of thermal image pixel points, infrared high-temperature anomaly area ratio (IHAR), is proposed to quantitatively analyze the infrared thermal image characteristics in combination with image processing technology. The results show that: after the denoising and enhancement based on the double-domain decomposition and improved CLAHE algorithm, the infrared thermal image effect is improved; the change of IHAR index is controlled by stress, with significant characteristics of stages, which are synchronized and consistent with the infrared thermal image sequence; The precursors of IHAR show significant increase and fluctuation, and the time point is around 60% to 65% of the peak stress; compared with other infrared indicators, IHAR is better in terms of stage, easy identification of precursors, and validity of damage characterization, which can effectively reflect the evolution process of coal loaded damage, and provide early warning information for coal destabilization monitoring.

Analysis of Adaptive Morphology for Visualization of Information Communication in Advertising Design

Abstract This paper uses adaptive morphology to construct an adaptive elliptic structure element that changes adaptively according to the image’s local neighborhood information. To calculate convolution, a Gaussian kernel with the outer product of the image gradient is used to achieve the diffusion of the linear matrix values of the initial matrix. The spatial distance of the pixels determines the size of the structural elements, and the pixel points that do not exceed the threshold are selected. The adaptive morphology method resulted in a 27.7 difficulty level for visual communication design, and the algorithm’s accuracy was 87.18%. Advertising design can break through dimensional limitations in visual space through adaptive morphology to generate interaction, indicating a path for development.

Design of automatic tobacco trash detection system based on quadratic fuzzy theory

Abstract Fuzzy sets and deep learning models have accurate recognition accuracy in automatic detection. To be able to accurately identify tobacco debris and improve the quality of cigarettes, this paper proposes the design of a tobacco debris detection system with a deep learning model and fuzzy quadratic theory. The collected data is stored to AL422B by quadratic fuzzy set timing conversion, the cached image data is read by the microcontroller, the image pixel points are output, and the external controller and internal registers are set to read the tobacco image. The integrated Thumb extended instruction set to obtain the CPU clock frequency makes the controller interrupt and ensures PWM output. Integrate the internal fixed oscillator and external integrated control circuit to debug the program interface circuit to prevent power on and off misoperation. The negative log-likelihood function is obtained by following the activation rules given by the visible layer and hidden layer activation functions. The RBM is trained by quadratic fuzzy set estimation to optimize the parameters; test sample sets appear overfitting, combined with DBN pre-training and fine-tuning, iterative output, and labeling target data to meet the preset requirements to achieve intelligent detection of tobacco debris. The result analysis shows that the deep learning model and the quadratic fuzzy set generalization ability and accuracy are high, the highest F-value in tobacco clutter detection reaches 100%, and the system is designed to detect tobacco clutter automatically with high accuracy and good detection of clutter.

Performance Evaluation of GPM IMERG Precipitation Products over the Tropical Oceans Using Buoys

Abstract The majority of global precipitation falls in tropical oceans. Nonetheless, due to the lack of in situ precipitation measurements, the number of studies over the tropical oceans remains limited. Similarly, the performance of IMERG products over the tropical oceans is unknown. In this context, this study quantitatively evaluates the 20 years (2001–20) of IMERG V06 Early, Late, and Final products against the in situ buoys’ estimates using the pixel–point approach at a daily scale across the tropical oceans. Results show that IMERG represents well the mean spatial pattern and spatial variation of precipitation, though significant differences exist in the magnitude of precipitation amount. Overall, IMERG notably overestimates precipitation across the tropical ocean, with maxima over the western Pacific and Indian Oceans, while it performs better over the eastern Pacific and Atlantic Oceans. Moreover, irrespective of the region, IMERG sufficiently detects precipitation events (i.e., >0.1 mm day−1) for high-precipitation regions, though it significantly overestimates the magnitude. Despite IMERG’s detection issues of precipitation events over the regions with lower precipitation, it is in good agreement with the buoys in total precipitation estimation. The positive hit bias and false alarm bias are the major contributions to the overall total positive bias. Furthermore, the detection capability of IMERG tends to decline with increasing precipitation rates. In terms of IMERG runs, the IMERG Final product performs slightly better than the Early and Late runs. More detailed studies over the tropical oceans are required to better characterize the biases and their sources.

Perceptual cue-guided adaptive image downscaling for enhanced semantic segmentation on large document images

AbstractImage downscaling is an essential operation to reduce spatial complexity for various applications and is becoming increasingly important due to the growing number of solutions that rely on memory-intensive approaches, such as applying deep convolutional neural networks to semantic segmentation tasks on large images. Although conventional content-independent image downscaling can efficiently reduce complexity, it is vulnerable to losing perceptual details, which are important to preserve. Alternatively, existing content-aware downscaling severely distorts spatial structure and is not effectively applicable for segmentation tasks involving document images. In this paper, we propose a novel image downscaling approach that combines the strengths of both content-independent and content-aware strategies. The approach limits the sampling space per the content-independent strategy, adaptively relocating such sampled pixel points, and amplifying their intensities based on the local gradient and texture via the content-aware strategy. To demonstrate its effectiveness, we plug our adaptive downscaling method into a deep learning-based document image segmentation pipeline and evaluate the performance improvement. We perform the evaluation on three publicly available historical newspaper digital collections with differences in quality and quantity, comparing our method with one widely used downscaling method, Lanczos. We further demonstrate the robustness of the proposed method by using three different training scenarios: stand-alone, image-pyramid, and augmentation. The results show that training a deep convolutional neural network using images generated by the proposed method outperforms Lanczos, which relies on only content-independent strategies.

Research on Identification and Detection of Transmission Line Insulator Defects Based on a Lightweight YOLOv5 Network

Transmission line fault detection using drones provides real-time assessment of the operational status of transmission equipment, and therefore it has immense importance in ensuring stable functioning of the transmission lines. Currently, identification of transmission line equipment relies predominantly on manual inspections that are susceptible to the influence of natural surroundings, resulting in sluggishness and a high rate of false detections. In view of this, in this study, we propose an insulator defect recognition algorithm based on a YOLOv5 model with a new lightweight network as the backbone network, combining noise reduction and target detection. First, we propose a new noise reduction algorithm, i.e., the adaptive neighborhood-weighted median filtering (NW-AMF) algorithm. This algorithm employs a weighted summation technique to determine the median value of the pixel point’s neighborhood, effectively filtering out noise from the captured aerial images. Consequently, this approach significantly mitigates the adverse effects of varying noise levels on target detection. Subsequently, the RepVGG lightweight network structure is improved to the newly proposed lightweight structure called RcpVGG-YOLOv5. This structure facilitates single-branch inference, multi-branch training, and branch normalization, thereby improving the quantization performance while simultaneously striking a balance between target detection accuracy and speed. Furthermore, we propose a new loss function, i.e., Focal EIOU, to replace the original CIOU loss function. This optimization incorporates a penalty on the edge length of the target frame, which improves the contribution of the high-quality target gradient. This modification effectively addresses the issue of imbalanced positive and negative samples for small targets, suppresses background positive samples, and ultimately enhances the accuracy of detection. Finally, to align more closely with real-world engineering applications, the dataset utilized in this study consists of machine patrol images captured by the Unmanned Aerial Systems (UAS) of the Yunnan Power Supply Bureau Company. The experimental findings demonstrate that the proposed algorithm yields notable improvements in accuracy and inference speed compared to YOLOv5s, YOLOv7, and YOLOv8. Specifically, the improved algorithm achieves a 3.7% increase in accuracy and a 48.2% enhancement in inference speed compared to those of YOLOv5s. Similarly, it achieves a 2.7% accuracy improvement and a 33.5% increase in inference speed compared to those of YOLOv7, as well as a 1.5% accuracy enhancement and a 13.1% improvement in inference speed compared to those of YOLOv8. These results validate the effectiveness of the proposed algorithm through ablation experiments. Consequently, the method presented in this paper exhibits practical applicability in the detection of aerial images of transmission lines within complex environments. In future research endeavors, it is recommended to continue collecting aerial images for continuous iterative training, to optimize the model further, and to conduct in-depth investigations into the challenges associated with detecting small targets. Such endeavors hold significant importance for the advancement of transmission line detection.

ARE-Net: An Improved Interactive Model for Accurate Building Extraction in High-Resolution Remote Sensing Imagery

Accurate building extraction for high-resolution remote sensing images is critical for topographic mapping, urban planning, and many other applications. Its main task is to label each pixel point as a building or non-building. Although deep-learning-based algorithms have significantly enhanced the accuracy of building extraction, fully automated methods for building extraction are limited by the requirement for a large number of annotated samples, resulting in a limited generalization ability, easy misclassification in complex remote sensing images, and higher costs due to the need for a large number of annotated samples. To address these challenges, this paper proposes an improved interactive building extraction model, ARE-Net, which adopts a deep interactive segmentation approach. In this paper, we present several key contributions. Firstly, an adaptive-radius encoding (ARE) module was designed to optimize the interaction features of clicks based on the varying shapes and distributions of buildings to provide maximum a priori information for building extraction. Secondly, a two-stage training strategy was proposed to enhance the convergence speed and efficiency of the segmentation process. Finally, some comprehensive experiments using two models of different sizes (HRNet18s+OCR and HRNet32+OCR) were conducted on the Inria and WHU building datasets. The results showed significant improvements over the current state-of-the-art method in terms of NoC90. The proposed method achieved performance enhancements of 7.98% and 13.03% with HRNet18s+OCR and 7.34% and 15.49% with HRNet32+OCR on the WHU and Inria datasets, respectively. Furthermore, the experiments demonstrated that the proposed ARE-Net method significantly reduced the annotation costs while improving the convergence speed and generalization performance.

Dual-Core Adaptive NLM Image Denoising Algorithm Based on Variable-Size Window and Neighborhood Multifeatures

To solve the problem that the similarity calculation between neighbors was easily disturbed by noise in the traditional nonlocal mean (NLM) denoising algorithm, a dual-core NLM denoising algorithm based on neighborhood multifeatures and variable-size search window was proposed. The algorithm first proposed to use the eigenvalues of the structure tensor to classify the region where the target pixel points were located and used different sizes of the search window to search for similar neighborhoods for target pixel points in different categories of the region, thus effectively avoiding the problem of oversmoothing or inadequate denoising of the image caused by the use of the global size. Then, the gradient features between image blocks were defined and combined with grayscale features and spatial features to measure the similarity of neighborhood blocks, which solved the problem of noise interfering with the search of similar blocks. Then, an adaptive algorithm with Gaussian–Sinusoidal dual kernel function and quantitative estimation of the optimal values of the filtering parameters was designed to calculate the neighborhood similarity weights to improve the accuracy of image denoising. Finally, the similarity weights were used to weight and average the search neighborhood of the target pixel points to achieve the denoising of the target pixel points. To test the effectiveness of the algorithm, denoising tests were performed using multiple standard grayscale images with different levels of Gaussian white noise added and compared with several advanced denoising algorithms. The experimental results showed that the algorithm was effective. The algorithm improved the image peak signal-to-noise ratio by more than 56.54% on average when Gaussian white noise was removed, and the structural similarity reached more than 0.701 on average. Compared with the traditional NLM algorithm and other improved algorithms, the algorithm proposed in this paper had strong denoising ability, better protection of edges and texture details, and the quality of the image was greatly improved, which had a good application prospect.

Deep learning-based multiscale CNN-based U network model for leaf disease diagnosis and segmentation of lesions in tomato

BackgroundThe advent of deep learning technologies has expanded the utilization of convolutional neural networks in diagnosing and segmenting plant leaf diseases. However, the segmentation of lesions, especially those with diverse sizes and shapes, often requires extensive pixel-level annotations. ObjectiveThis study aims to introduce a multiscale U-network for effective concurrent segmentation and diagnosis of tomato leaf lesions. MethodsOur model employs multi-scale residual modules to adapt dynamically to diseases that alter lesion size and shape. The Classifier and Bridge (CB) module is introduced to connect the disease feature extraction stage with the disease spot segmentation stage, providing crucial information about a specific lesion type. During the segmentation stage, the activation map of a specific class is deconvoluted using a combination of up-sampling and convolution. This deconvolution feature is fused with the low-level feature through a skip connection strategy. A limited number of pixel-level labels are used, and the binary cross-entropy loss function is applied for supervised training at each pixel point, directing the feature extraction network to focus on the lesion location. ResultsThe model was evaluated using an original test set and an interference test set simulating noise and light intensity changes. The multiscale U-network demonstrated an average pixel accuracy of approximately 99.2%, a resilience to brightness reduction by about 92.4%, and an ability to handle up to 0.98 salt and pepper noise at a rate of about 99.2%, showcasing its robustness against various image alterations. The GFLOPs for the proposed method was found to be approximately 3.13e+06, indicating its computational efficiency. ConclusionThe experimental results highlight the potential of the proposed multiscale U-network in effectively segmenting and identifying lesions in tomato leaf tissue. These findings suggest promising opportunities for applying this model in the broader field of agricultural disease management.

Efficient railway track region segmentation algorithm based on lightweight neural network and cross-fusion decoder

To segment railway track regions in real-time for intrusion detection and improving security, this paper proposes an efficient railway track region segmentation network (ERTNet) based on the encoder–decoder architecture. Firstly, to ensure the lightweight of the encoder, depthwise convolution and the channel shuffle are utilized to construct sandglass-type feature extraction unit. Secondly, a feature-matching-based cross-fusion decoder is utilized to fuse deep and shallow feature maps. Thirdly, the knowledge distillation is employed with large-scale Deeplab v3+ as the teacher model to improve performance. Additionally, a loss function is proposed to penalize pixel points with large offsets. Finally, the ERTNet is validated on the self-built dataset, achieving an MIoU (Mean Intersection over Union) of 92.4% , which is 5.22% improvement over the benchmark model. ERTNet achieves a balance between segmentation accuracy and computational efficiency, requiring only 0.5 M parameters and 0.92 G FLOPs (Floating Point Operations).