Search Results (24)

This paper presents an analysis of the impact of the control architecture in a mobile robot on the quality of regulation in real systems. Comparative studies were conducted for successive stages of the implementation of architectural improvements, such as optimization of RTOS resource utilization, the use of hardware mechanisms (DMP, DMA) for sensor data acquisition, and the optimization of subordinate controllers. The results showed that the final control quality depends not only on the controller tuning but also on the efficient management of the hardware and software resources of the control system. Retuning the PID controller after architectural modifications enabled the achievement of a better control quality index (IAE). The novelty of this work lies in demonstrating, through experimental evaluation, that embedded control architecture has a measurable and systematic impact on regulation quality in real systems. The obtained results indicate a significant relationship between control architecture and control performance, representing an important step toward bridging the gap between simulation studies and real-world implementations in mobile robotics. Full article

To improve traction efficiency and reduce fuel consumption during tractor tillage operations, a coordinated control method for electro-hydraulic hitch depth and tractor speed was proposed. Based on theoretical analysis, a dynamic model of the tractor–implement system during tillage was established. The principles of coordinated control were developed, and a comprehensive performance evaluation index considering both efficiency and economic performance of the tractor was proposed to optimize the coordinated control objectives. A depth controller and a speed controller were, respectively, designed based on the sliding mode control algorithm. A hardware-in-the-loop test platform for coordinated control of electro-hydraulic hitch depth and travel speed was established via CAN communication. Comparative experiments were conducted under three operational conditions (Condition 1: tillage depth 16 cm, soil specific resistance 2.5–3.5 N/cm²; Condition 2: 20 cm, 3.5–4.5 N/cm²; Condition 3: 24 cm, 4.5–5.5 N/cm²) against the conventional single depth control method under full throttle. Results demonstrated that the coordinated depth-speed control method improved the overall tractor efficiency-economy by 50.0%, 33.3%, and 26.7% under these respective conditions compared to the single depth control method. This method not only ensures operation quality but also enhances the comprehensive performance of the tractor, effectively improving traction efficiency and reducing fuel consumption. Moreover, it demonstrates better adaptability to varying field conditions. Full article

Digitally controlled DC–DC converters are vulnerable to sensor-side spoofing, motivating plant-level anomaly detection that respects the converter physics. We present a physics-informed LSTM (PI–LSTM) autoencoder for a 24→12 V buck converter. The model embeds discrete-time circuit equations as residual penalties and uses a fixed decision rule ($\tau =\mu +3\sigma$, $N=3$ consecutive samples). We study three voltage-sensing attacks (DC bias, fixed-sample delay, and narrowband noise) in MATLAB/Simulink. We then validate the detection path on a TMS320F28379 DSP. The detector attains F1 scores of 96.12%, 91.91%, and 97.50% for bias, delay, and noise (simulation); on hardware, it achieves 2.9–4.2 ms latency with an alarm-wise FPR of ≤1.2%. We also define a unified safety box for DC rail quality and regulation. In simulations, we evaluate a confusion index (CI) policy for safety-bounded performance adjustment. A operating point yields $\mathrm{CI}\approx 0.25$ while remaining within the safety limits. In hardware experiments without CI actuation, the $V_{r,pp}$ and IRR stayed within the limits, whereas the ±2% regulation window was occasionally exceeded under the delay attack (up to ≈2.8%). These results indicate that physics-informed detection is deployable on resource-constrained controllers with millisecond-scale latency and a low alarm-wise FPR, while the full hardware validation of CI-guided deception (safety-bounded performance adjustment) under the complete safety box is left to future work. Full article

High Dynamic Range (HDR) imaging is capable of capturing vivid and lifelike visual effects, which are crucial for fields such as computer vision, photography, and medical imaging. However, real-time processing of HDR content remains challenging due to the computational complexity of tone mapping algorithms and the inherent limitations of Low Dynamic Range (LDR) capture systems. This paper presents an adaptive HDR tone mapping processor that achieves high computational efficiency and robust image quality under varying exposure conditions. By integrating an exposure-adaptive factor into a bilateral filtering framework, we dynamically optimize parameters to achieve consistent performance across fluctuating illumination conditions. Further, we introduce a high-radix Taylor expansion technique to accelerate floating-point logarithmic and exponential operations, significantly reducing resource overhead while maintaining precision. The proposed architecture, implemented on a Xilinx XCVU9P FPGA, operates at 250 MHz and processes 4K video at 30 frames per second (FPS), outperforming state-of-the-art designs in both throughput and hardware efficiency. Experimental results demonstrate superior image fidelity with an average Tone Mapping Quality Index (TMQI): 0.9314 and 43% fewer logic resources compared to existing solutions, enabling real-time HDR processing for high-resolution applications. Full article

Semantic segmentation of medical images holds significant potential for enhancing diagnostic and surgical procedures. Radiology specialists can benefit from automated segmentation tools that facilitate identifying and isolating regions of interest in medical scans. Nevertheless, to obtain precise results, sophisticated deep learning models tailored to this specific task must be developed and trained, a capability not universally accessible. Segment Anything Model (SAM) 2 is a foundational model designed for image and video segmentation tasks, built on its predecessor, SAM. This paper introduces a novel approach leveraging SAM 2’s video segmentation capabilities to reduce the prompts required to segment an entire volume of medical images. The study first compares SAM and SAM 2’s performance in medical image segmentation. Evaluation metrics such as the Jaccard index and Dice score are used to measure precision and segmentation quality. Then, our novel approach is introduced. Statistical tests include comparing precision gains and computational efficiency, focusing on the trade-off between resource use and segmentation time. The results show that SAM 2 achieves an average improvement of 1.76% in the Jaccard index and 1.49% in the Dice score compared to SAM, albeit with a ten-fold increase in segmentation time. Our novel approach to segmentation reduces the number of prompts needed to segment a volume of medical images by 99.95%. We demonstrate that it is possible to segment all the slices of a volume and, even more, of a whole dataset, with a single prompt, achieving results comparable to those obtained by state-of-the-art models explicitly trained for this task. Our approach simplifies the segmentation process, allowing specialists to devote more time to other tasks. The hardware and personnel requirements to obtain these results are much lower than those needed to train a deep learning model from scratch or to modify the behavior of an existing one using model modification techniques. Full article

In the present context, the asymmetrical multilevel inverter (AMLI) with minimal switches is extensively used to achieve high power quality across a load. The conventional inverter suffers from the disadvantages of large system components, low efficiency, high THD, high losses, etc. The asymmetrical multilevel inverter therefore offers an appropriate alternative to address the key issues of conventional inverters. The present paper aims to analyze the performance comparison of the proposed 15-level AMLI structure with different switching techniques, i.e., NLC, SHEPWM and SPWM. Moreover, loss analysis of the considered AMLI has been performed for different switching techniques. Furthermore, different performance parameters such as conduction losses, switching losses, total losses, THD, efficiency and power delivery of the inverter have been evaluated. It has been observed that the considered inverter topology offers the superior performance with the SHEPWM modulation technique at a modulation index of 0.8. Finally, hardware arrangement of the inverter has also been developed in the laboratory with a real-time Opal-RT 4510 simulator to verify the accuracy of simulation results. Full article

Super-resolution (SR) techniques have shown significant promise in enhancing the resolution of MRI images, which are often limited by hardware constraints and acquisition time. In this study, we introduce an advanced regularization method for MRI super-resolution that integrates spatially adaptive techniques with a robust denoising process to improve image quality. The proposed method excels in preserving high-frequency details while effectively suppressing noise, addressing common limitations of conventional SR approaches. The validation of clinical MRI datasets demonstrates that our approach achieves superior performance compared to traditional algorithms, yielding enhanced image clarity and quantitative improvements in metrics such as the peak signal-to-noise ratio (PSNR) and structural similarity index (SSIM). Full article

This paper presents the operation of a modified speed controller with a standard PI/PID structure that includes the preprocessing of the controller’s input signal, focusing on the past behavior of control errors. The modification involves adding a delay line, with the outputs of the individual line segments summed with a weighting method, as detailed in the paper. One of the significant advantages of this method is its use of a standard industrial controller structure, which makes it highly practical and easily implementable in existing systems. By relying on well-established control frameworks, this approach reduces the need for specialized hardware or complex modifications, allowing for smoother integration and lower implementation costs. The delay-based signal shaping shows excellent properties for the electric drive system powered by a hard-switching PWM converter. The set of weighted delays acts as a filter whose parameters are chosen using the quality function to test different configurations for optimal performance. When tested in a speed control system for a Permanent Magnet Synchronous Motor, the modifications improved the control quality index, indicating better performance and efficiency. Importantly, the system allows for reducing or eliminating the gain in the differentiating part of the controller, which decreases motor current chattering and noise. This paper includes an experimental verification of the proposed solution in a laboratory setting under semi-industrial conditions. Full article

The printed circuit board (PCB) winding coil offers advantages such as small size, high precision, high repeatability, and low cost, making it conducive to the miniaturization of electronic equipment and a popular choice in wireless power transmission systems. This paper aims to clarify the correlation between induction parameters and inductive capabilities using the orthogonal array of the modified Taguchi method for Pedelec applications. The conventional Taguchi method typically achieves only local optimization; however, this paper considers practical application conditions and combines experimental data to establish the initial values of the orthogonal array, thereby achieving global optimization. Additionally, the tuning process of the Taguchi method replaces physical experiments with simulations, enhancing optimization speed and reducing hardware implementation costs. The performance index for the proposed modified Taguchi tuning method is selected as a combination of the quality factor (Q) and coupling coefficient (k) to minimize AC resistance and improve system efficiency. To validate the proposed method, the designed coils were implemented and tested in a WPT system based on S–S compensation with a half-bridge topology. The experimental results demonstrate that the optimized PCB coil parameters derived from the proposed tuning method accurately validate the method’s effectiveness and accuracy. From the measured results with the proposed modified tuning method, the system efficiency is increased by 43.87% and the system transmitting power is increased by 28.51%. Full article

This article aims to assess the usability of selected map portals with a checklist. The methods employed allowed the author to conduct user experience tests from a longer temporal perspective against a retrospective analysis of the evolution of design techniques for presenting spatial data online. The author performed user experience tests on three versions of Tomice Municipality’s geoportal available on the Internet. The desktop and mobile laboratory tests were performed by fourteen experts following a test scenario. The study employs the exploratory approach, inspection method, and System Usability Scale (SUS). The author calculated the Geoportal Overall Quality (GOQ) index to better illustrate the relationships among the subjective perceptions of the usability quality of the three geoportals. The usability results were juxtaposed with performance measurements. Normalised and aggregated results of user experience demonstrated that the expert assessments of the usability of geoportals G1 and G3 on mobile devices were similar despite significant development differences. The overall results under the employed research design have confirmed that geoportal G2 offers the lowest usability in both mobile and desktop modes. The study has demonstrated that some websites can retain usability even considering the dynamic advances in hardware and software despite their design, which is perceived as outdated today. Users still expect well-performing and quick map applications, even if this means limited functionality and usability. Moreover, the results indirectly show that the past resolution of the ‘large raster problem’ led to the aggravation of the issue of ‘large scripts’. Full article

The popularity of handheld devices for point-of-care ultrasound (POCUS) has increased in recent years due to their portability and cost-effectiveness. However, POCUS has the drawback of lower imaging quality compared to conventional ultrasound because of hardware limitations. Improving the quality of POCUS through post-image processing would therefore be beneficial, with deep learning approaches showing promise in this regard. This review investigates the state-of-the-art progress of image enhancement using deep learning suitable for POCUS applications. A systematic search was conducted from January 2024 to February 2024 on PubMed and Scopus. From the 457 articles that were found, the full text was retrieved for 69 articles. From this selection, 15 articles were identified addressing multiple quality enhancement aspects. A disparity in the baseline performance of the low-quality input images was seen across these studies, ranging between 8.65 and 29.24 dB for the Peak Signal-to-Noise Ratio (PSNR) and between 0.03 an 0.71 for the Structural Similarity Index Measure (SSIM). In six studies, where both the PSNR and the SSIM metrics were reported for the baseline and the generated images, mean differences of 6.60 (SD ± 2.99) and 0.28 (SD ± 0.15) were observed for the PSNR and SSIM, respectively. The reported performance outcomes demonstrate the potential of deep learning-based image enhancement for POCUS. However, variability in the extent of the performance gain across datasets and articles was notable, and the heterogeneity across articles makes quantifying the exact improvements challenging. Full article

Background: Diffusion-weighted imaging (DWI), a pivotal component of multiparametric magnetic resonance imaging (mpMRI), plays a pivotal role in the detection, diagnosis, and evaluation of gastric cancer. Despite its potential, DWI is often marred by substantial anatomical distortions and sensitivity artifacts, which can hinder its practical utility. Presently, enhancing DWI’s image quality necessitates reliance on cutting-edge hardware and extended scanning durations. The development of a rapid technique that optimally balances shortened acquisition time with improved image quality would have substantial clinical relevance. Objectives: This study aims to construct and evaluate the unsupervised learning framework called attention dual contrast vision transformer cyclegan (ADCVCGAN) for enhancing image quality and reducing scanning time in gastric DWI. Methods: The ADCVCGAN framework, proposed in this study, employs high b-value DWI (b = 1200 ${\mathrm{s}/\mathrm{mm}}^2$) as a reference for generating synthetic b-value DWI (s-DWI) from acquired lower b-value DWI (a-DWI, b = 800 ${\mathrm{s}/\mathrm{mm}}^2$). Specifically, ADCVCGAN incorporates an attention mechanism CBAM module into the CycleGAN generator to enhance feature extraction from the input a-DWI in both the channel and spatial dimensions. Subsequently, a vision transformer module, based on the U-net framework, is introduced to refine detailed features, aiming to produce s-DWI with image quality comparable to that of b-DWI. Finally, images from the source domain are added as negative samples to the discriminator, encouraging the discriminator to steer the generator towards synthesizing images distant from the source domain in the latent space, with the goal of generating more realistic s-DWI. The image quality of the s-DWI is quantitatively assessed using metrics such as the peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), feature similarity index (FSIM), mean squared error (MSE), weighted peak signal-to-noise ratio (WPSNR), and weighted mean squared error (WMSE). Subjective evaluations of different DWI images were conducted using the Wilcoxon signed-rank test. The reproducibility and consistency of b-ADC and s-ADC, calculated from b-DWI and s-DWI, respectively, were assessed using the intraclass correlation coefficient (ICC). A statistical significance level of p < 0.05 was considered. Results: The s-DWI generated by the unsupervised learning framework ADCVCGAN scored significantly higher than a-DWI in quantitative metrics such as PSNR, SSIM, FSIM, MSE, WPSNR, and WMSE, with statistical significance (p < 0.001). This performance is comparable to the optimal level achieved by the latest synthetic algorithms. Subjective scores for lesion visibility, image anatomical details, image distortion, and overall image quality were significantly higher for s-DWI and b-DWI compared to a-DWI (p < 0.001). At the same time, there was no significant difference between the scores of s-DWI and b-DWI (p > 0.05). The consistency of b-ADC and s-ADC readings was comparable among different readers (ICC: b-ADC 0.87–0.90; s-ADC 0.88–0.89, respectively). The repeatability of b-ADC and s-ADC readings by the same reader was also comparable (Reader1 ICC: b-ADC 0.85–0.86, s-ADC 0.85–0.93; Reader2 ICC: b-ADC 0.86–0.87, s-ADC 0.89–0.92, respectively). Conclusions: ADCVCGAN shows excellent promise in generating gastric cancer DWI images. It effectively reduces scanning time, improves image quality, and ensures the authenticity of s-DWI images and their s-ADC values, thus providing a basis for assisting clinical decision making. Full article

The current sorting process for winter jujubes relies heavily on manual labor, lacks uniform sorting standards, and is inefficient. Furthermore, existing devices have simple structures and can only be sorted based on size. This paper introduces a method for detecting surface defects on winter jujubes using convolutional neural networks (CNNs). According to the current situation in the winter jujube industry in Zhanhua District, Binzhou City, Shandong Province, China, we collected winter jujubes with different surface qualities in Zhanhua District; produced a winter jujube dataset containing 2000 winter jujube images; improved it based on the traditional AlexNet model; selected a total of four classical convolutional neural networks, AlexNet, VGG-16, Inception-V3, and ResNet-34, to conduct different learning rate comparison training experiments; and then took the accuracy rate, loss value, and F1-score of the validation set as evaluation indexes while analyzing and discussing the training results of each model. The experimental results show that the improved AlexNet model had the highest accuracy in the binary classification case, with an accuracy of 98% on the validation set; the accuracy of the Inception V3 model reached 97%. In the detailed classification case, the accuracy of the Inception V3 model was 95%. Different models have different performances and different hardware requirements, and different models can be used to build the system according to different needs. This study can provide a theoretical basis and technical reference for researching and developing winter jujube detection devices. Full article

This study rigorously investigates the effectiveness of nonlinear filters in CMOS for 2-D signal processing to enhance image quality. We comprehensively compare traditional linear filters’ performance, which operate on the principle of linearity, with nonlinear filters, such as the median-median (Med-Med) approach, designed to handle nonlinear data. To ensure the validity of our findings, we use widely accepted metrics like normalized squared error (NSE), peak signal-to-noise ratio (PSNR), and structural similarity index (SSIM) to quantify the differences. Our simulations and experiments, conducted under controlled conditions, demonstrate that nonlinear filters in CMOS outperform linear filters in removing impulse noise and enhancing images. We also address the challenges of implementing these algorithms at the hardware level, focusing on power consumption and chip area optimization. Additionally, we propose a new architecture for the Med-Med filter and validate its functionality through experiments using a 9-pixel image sensor array. Our findings highlight the potential of nonlinear filters in CMOS for real-time image quality enhancement and their applicability in various real-world imaging applications. This research contributes to visual technology by combining theoretical insights with practical implementations, paving the way for more efficient and adaptable imaging systems. Full article

Ultrasound is extremely efficient for wireless signal transmission through metal barriers due to no limit of the Faraday shielding effect. Echoing in the ultrasonic channel is one of the most challenging obstacles to performing high-quality communication, which is generally coped with by using a channel equalizer or pre-distorting filter. In this study, a deep learning algorithm called a dual-path recurrent neural network (DPRNN) was investigated for echo cancellation in an ultrasonic through-metal communication system. The actual system was constructed based on the combination of software and hardware, consisting of a pair of ultrasonic transducers, an FPGA module, some lab-made circuits, etc. The approach of DPRNN echo cancellation was applied to signals with a different signal-to-noise ratio (SNR) at a 2 Mbps transmission rate, achieving higher than 20 dB SNR improvement for all situations. Furthermore, this approach was successfully used for image transmission through a 50 mm thick aluminum plate, exhibiting a 24.8 dB peak-signal-to-noise ratio (PSNR) and a about 95% structural similarity index measure (SSIM). Additionally, compared with three other echo cancellation methods—LMS, RLS and PNLMS—DPRNN has demonstrated higher efficiency. All those results firmly validate that the DPRNN algorithm is a powerful tool to conduct echo cancellation and enhance the performance of ultrasonic through-metal transmission. Full article