International Journal of Engineering and Manufacturing @ijem
Journal articles - International Journal of Engineering and Manufacturing
All articles: 649
A Checkpointing Algorithm Based Unreliable Non-FIFO Channels
Scientific article
We propose a coordinated checkpointing algorithm based unreliable non-FIFO channel. In unreliable non-FIFO channel, the system can lose, duplicate, or reorder messages. The processes may not compute some messages because of message losses; the processes may compute some messages twice or more because of message duplicate; the processes may not compute messages according to their sending order because of message reordering. The above-mentioned problems make processes produce incorrect computation result, consequently, prevent processes from taking consistent global checkpoints. Our algorithm assigns each message a sequence number in order to resolve above-mentioned problems. During the establishing of the checkpoint, the consistency of checkpoint can be determined by the sequence number of sending and receiving messages. We can identify the lost messages, reordering messages and duplicate messages by checking the sequence number of sending and receiving messages. We resolve above-mentioned problems by resending the lost messages, buffering the reordering messages and dropping the duplicate messages. Our algorithm makes processes take consistent global checkpoints.
Free
Scientific article
Accurate air quality mapping in regions with sparse sensor deployment remains a challenge due to high infrastructure costs. While modern literature increasingly favors heavy, cloud-based Artificial Intelligence frameworks that assume dense input networks, the operational boundaries and mathematical fidelity of lean, edge-computed spatial interpolation models in ultra-sparse (e.g., 5-node) live frameworks remain poorly defined. This work presents a case-study comparative evaluation of conventional spatial interpolation techniques for estimating pollutant concentrations in sensor-limited environments. Eight interpolation methods, namely Inverse Distance Weighting (IDW), Kriging, Empirical Bayesian Kriging (EBK), Radial Basis Function (RBF), Nearest Neighbor (NN), Akima, Piecewise Cubic Hermite Interpolation (PCHIP), and Cubic Spline (CS), were analyzed using real-time environmental data collected from five locations in Kalady, Kerala, India. The interpolation performance was evaluated using RMSE and MAE metrics through leave-one-out validation. Results indicate that, within this deployment, Kriging and IDW achieved the best estimation accuracy for PM2.5, PM10, humidity, and temperature compared to the other methods evaluated. A LoRa-based sensing framework was also integrated to support low-power real-time environmental monitoring. As a single-region case study based on five monitoring nodes, the findings are specific to this deployment, and broader validation across additional stations and regions is identified as future work. The proposed approach demonstrates the feasibility of resource-efficient air quality estimation in regions with limited monitoring infrastructure.
Free
A Comparative Analysis of Video Summarization Techniques
Scientific article
Video summarization special field of signal processing which includes pre-processing of video sets, their contextual segmentation, application-specific feature extraction & selection, and identification of dissimilar frame sets. Various variety of machine learning models are proposed by researchers to design such summarization methods, and each of them varies in terms of their functional nuances, application-specific advantages, deployment specific limitations, and contextual future scopes. Moreover, these models also vary in terms of quantitative & qualitative measures including accuracy of summarization, computational complexity, delay needed for summarization, precision during the summarization process, etc. Due to such a wide variation in performance levels, it is difficult for researchers to identify optimal models for their functional-specific &performance-specific use cases. Because of this, researchers and summarization-system-designers are required to validate individual models, which increases the delay & cost needed for final model deployments. To overcome these delays & reduce deployment costs, this paper initially discusses a multiple variety of video summarization models in terms of their working characteristics. Based on this discussion, researchers shall be able to identify optimum models for their functionality-specific use cases. This paper also analyzes and compares the reviewed models in terms of their performance metrics including summarization accuracy, delay, complexity, scalability and fMeasure, which will further allow readers to identify performance-specific models for their deployments. A novel Summarization Rank Metric (SRM) is calculated based on these evaluation metrics, which will assist readers to identify models that can perform optimally w.r.t. multiple evaluation parameters & different use cases. This metric is calculated by combining all the comparison metrics, which will assist in identification of models that have high accuracy, low delay, low complexity, high scalability & fMeasure levels.
Free
Scientific article
In recent years, the rising prevalence of chronic illness has led to an increase in disability of patients. Extensive research has been done to enhance both the functional abilities as well as the quality of the affected individuals’ lives. Researchers have worked on the effects of numerous scholars, keywords and countries of these specific fields. However, a few state-of-the-art bibliometric analyses have been done in this research to reduce the quantitative aspects of the vast research fields of rehabilitation. We have precisely selected 427 core papers from the Web of Science database spanning from 1999 to 2022 where Machine Learning (ML) or Deep Learning (DL) is used in the rehabilitation field. Consequently, our analysis focuses on citation patterns, trend analysis and collaborations between countries or influential keywords offering a detailed overview of global trends in this interdisciplinary domain. Additionally, we visualize the research trends of various authors and countries which provide invaluable insights into research impact as well as collaboration networks. Overall, this paper aims to shape the evolving field of rehabilitation by providing in depth analysis of the citation landscape, key researchers, and international collaborations.
Free
A Discriminative Statistical Model for Digital Image Forgery Detection
Scientific article
The headway of modern technology and facility to use processing software leads to tamper and implicate of digital images. This tampering is being performed without leaving any a clear effect noted with the naked eye. The discrimination between different authentic and forged images can be based on its Probability Density Functions (PDFs). This paper introduces a new model for digital image forgery detection. This framework has two main phases; training and testing. In the training phase, the peak is calculated for the derivatives histogram of the illumination components by using homomorphic filter to separate the illumination components on each image. Firstly, the derivative of illumination histogram for authentic and forged images is calculated then the PDFs are estimated for authentic and forged images, finally the threshold is determined. In the testing phase, the determined threshold is tested with realistic dataset followed by using the selected bins for feature calculation in the prediction process. In the final prediction step, a detection and decision process is performed to obtain performance of the new model. This new model is provided a very effective performance. Different color image contrast systems RGB and HIS are studied and utilized for testing our model and compare between each channel for two systems to estimate performance and obtain more sensitive channel.
Free
A Federated Learning-based Hybrid compression Technique for 3D Medical Images
Scientific article
The use of multimedia communication has grown significantly in recent years, which has raised demand for image data compression. One popular technique for representing an image in an efficient format is image compression. It results in low rates of transmission by precisely lowering the number of bits required to store the images. Since medical image data is growing so quickly, there is a lot of research being done on how to upload and store large amounts of medical images in real time while having a limited amount of storage space and network bandwidth. But still, at this time, medical image compression technology is unable to optimize both rate and distortion. The goal of the proposed hybrid compression technique is to increase compression performance without losing the standard of the image. Even though they need a lot of storage, 3D medical images provide detailed information about disease. Optimal Multi-linear Singular Value Decomposition (OMLSVD) and deep auto-encoders are used in the current work to compress 3D healthcare images. The Federated Learning technique addresses the issues of data privacy and leakage by having each user train a model on its dataset before sending the model's local weights to a global federated server. So, use a federated server to get a global dataset weight without leaking or publishing datasets, protecting privacy. The quality of 3D compression images can be improved by using the proposed Hybrid approach. Experimental evaluation demonstrates SSIM values near 1 and high PSNR, indicating excellent reconstruction quality. Additionally, it compares the image with the compressed JPEG2000 and the proposed Hybrid approach. Since the images have different storage sizes, they all appear to be identical.
Free
A Fuzzy Programming Technique for Solving Multi-objective Structural Problem
Scientific article
This paper proposes a new fuzzy multi-objective optimization approach to solve a multi-objective nonlinear programming problem in context of a structural design. We have been developed a multi-objective structural problem of a planar truss structural model in fuzzy environment. Here, the objectives are (i) to minimize weight of the structure and (ii) to minimize the vertical deflection at loading point. In this model, the design variables are the cross-section of the truss members and the constraints are the stresses in members. This approach is used to solve the structural model under uncertainty based on different operator. A numerical illustration is given to support our approach.
Free
A General Simulation Framework Based on CAN Bus for Satellite Design
Scientific article
The running status of a satellite can be imitated realistically through the use of an interactive simulation system, in which actual hardware can join for testing and validation. To solve the problem of interfaces and communications between different members in a satellite simulation system, an on-board interactive simulation system based on CAN bus is established. Several simulation members are built according to the composition of real system. Interfaces of drivers included in simulation members are packaged into a uniform interface, thus the on-board simulation system is established with the foundation of CAN bus simulation framework. Uniform communication interface of each simulation member is designed and realized, the framework of simulation flow is established, and foundation of intercommunion and operation between simulation members is laid. Experiment results prove that the framework can work stably with great efficiency and high flexibility, and expected effect is achieved.
Free
Scientific article
The paper proposes a generalized method for constructing graph-logical models of the failure behavior of fault-tolerant multiprocessor systems. Such models are used for evaluating system reliability by means of statistical experiments based on simulation of failure behavior. The method is applicable to non-basic systems whose failure behavior cannot be characterized solely by the number of failed components and therefore requires more flexible modeling approaches. The proposed method is based on combining auxiliary models corresponding to different operating conditions of the system into a single model that correctly represents the overall failure behavior. In contrast to existing approaches, it imposes no restrictions on the graph structures of the auxiliary models and does not depend on the specific procedures used for their construction. The key idea of the approach is to integrate such models under a set of mutually exclusive logical conditions, each of which determines the applicability of a particular auxiliary model for a given system state. A set of model transformations is introduced, and it is shown that these transformations preserve model equivalence, that is, correspondence to the same failure behavior of the system. It is demonstrated that these transformations are sufficient to transform auxiliary models to graphs with identical structures, which is a necessary condition for their combination within the proposed framework. Several illustrative examples of the application of the method are provided. The correctness of the constructed models is validated through analysis of representative system states and through exhaustive evaluation over all possible states. The results for the considered example indicate that the proposed method can construct models that accurately represent system failure behavior while yielding more compact graph structures compared to the existing approach. At the same time, comparable logical complexity is maintained. The evaluation is limited to a representative example and exhaustive analysis of system states, and further validation on a broader class of systems remains a direction for future work.
Free
A Graph Learning Framework for Analyzing Smart Assistive Sensor Data in Elderly Fall Prediction
Scientific article
Falls among older adults represent a critical public health challenge, with approximately 37.3 million fall-related incidents reported globally each year. Early and accurate prediction of falls is essential to enable timely, proactive interventions and to reduce associated injuries and fatalities. This work introduces a graph-based machine learning framework that leverages data from the cStick, a smart assistive Internet of Medical Things (IoMT) device. Bipartite graphs are constructed to model static correlations between multivariate sensor inputs— including heart rate variability (HRV), pressure, distance, SpO2, blood sugar levels, and accelerometer readings— and fall outcomes encoded as no fall, predicted fall, or definite fall. SHAP (SHapley Additive exPlanations) values are further integrated to enhance model interpretability and to identify the most influential sensor features through feature-only graph projections. Kernel Density Estimation (KDE) plots and pairplots are used to visualize feature distributions across fall categories. The proposed framework demonstrates that Pressure and Distance exhibit the strongest correlations with fall decisions (1.000 and −0.946, respectively), providing actionable insights for risk stratification. The integration of graph-based analysis with SHAP interpretability improves both predictive accuracy and transparency, facilitating proactive interventions and enhancing the safety, autonomy, and well-being of elderly individuals in real-world assistive care settings.
Free
A Heterogeneous Agent-based Asset Pricing Model and Simulation
Scientific article
Behavioral finance models can help to explain several stylized facts in financial markets. As one of the behavioral financial theory, prospect theory describes how ‘irrational investors’ making decisions under uncertainty. In this paper, we present a heterogeneous agent-based asset pricing model, where parts of investors determine their demand for risky asset using prospect theory utility function. Time series generated from simulation show many stylized facts that can be observed in actual financial markets, such as abnormal distribution of asset returns, volatility clustering and equity premium. We also find that positive correlation between investors’ performance and their market share, negative correlation between investors’ performance and the loss aversion coefficient under certain market condition.
Free
A Hybrid Approach for Real-time Vehicle Monitoring System
Scientific article
In today's modern era, with the significant increase in the number of vehicles on the roads, there is a pressing need for an advanced and efficient system to monitor them effectively. Such a system not only helps minimize the chances of any faults but also facilitates human intervention when required. Our proposed method focuses on detecting vehicles through background subtraction, which leverages the benefits of various techniques to create a comprehensive vehicle monitoring solution. In general, when it comes to surveillance and monitoring moving objects, the initial step involves detecting and tracking these objects. For vehicle segmentation, we employ background subtraction, a technique that distinguishes foreground objects from the background. To target the most prominent regions in video sequences, our method utilizes a combination of morphological techniques. The advancements in vision-related technologies have proven to be instrumental in object detection and image classification, making them valuable tools for monitoring moving vehicles. Methods based on moving object detection play a vital role in real-time extraction of vehicles from surveillance videos captured by street cameras. These methods also involve the removal of background information while filtering out noisy data. In our study, we employ background subtraction-based techniques that continuously update the background image to ensure efficient output. By adopting this approach, we enhance the overall performance of vehicle detection and monitoring.
Free
A LL Subband Based Digital Watermarking in DWT
Scientific article
Digital watermarking is a technique which allows an individual to add hidden copyright notices or other verification messages to digital audio, video, image and documents in such a way that the watermark cannot be removed against different attacks. Digital Watermarking is a process to prevent the unauthorized access and modification of data. It ensures that the communication between two parties' remains secure. Digital watermarking can be performed in spatial and frequency domain. In the spatial domain, the watermark is embedded in the very existence of the pixel. In frequency domain, the transformation of any kind is applied and then information is embedded. In this paper, we proposed an approach to watermarking in frequency domain using DWT technique. The gray scale host image is divided into four sub bands: LL, HL, LH, HH and the watermark are inserted in the LL sub band using DWT technique. As the image is divided into four sub bands, a watermark of equal size of the LL sub band is inserted and the results are analyzed on the bases of different parameters such as PSNR and MSE. LL represents the average component of the host image which contains the maximum information of the image. In this approach the watermark is inserted in LL sub band using XOR operation. As the table IV shows, the imperceptibility of this method is quite good. Also, it shows good results when compared with existing methods.
Free
Scientific article
Plant disease detection is vital for agricultural sustainability and food security. While Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs) have achieved high accuracy in this domain, CNNs often require millions of parameters and substantial computation. ViTs suffer from the quadratic time and space complexity of self-attention (SA), limiting their use on resource-constrained devices. Although SA is capable of modelling long-range dependencies when symptoms are dispersed, many plant diseases exhibit small, localized lesions or texture changes; therefore, Neighborhood Attention (NA) offers a more efficient and targeted alternative by focusing on nearby regions rather than the entire image. This work proposes a custom Localized NA block implemented in TensorFlow/Keras that operates directly on CNN feature maps, bypassing patch embedding and transformer modules. A lightweight CNN is then developed by combining depth-wise separable convolutions with the proposed localized NA block. In addition, a 100-category plant disease dataset covering 16 crops is presented. The dataset is curated, class-balanced, and made publicly available to support reproducibility and encourage further research. The proposed 9-layer CNN, with just 1.7M parameters and a size of 6.74 MB, achieved a favorable balance between accuracy, model size, and computational efficiency, compared with MobileNetV1, MobileNetV2, DenseNet121, InceptionV3, MobileViT-XXS, and EfficientViT-M0, achieving 98.97%± 0.33% accuracy on PlantVillage and 93.36%± 0.28% on the proposed dataset. The ablation study showed that the NA block improved test accuracy by approximately 2–3%, while Grad-CAM visualizations indicated more precise targeting of diseased areas in the leaf image.
Free
Scientific article
Face recognition is widely used for biometric authentication in applications such as mobile devices, financial services, intelligent surveillance, access control, and border security. However, face recognition systems remain vulnerable to presentation attacks, including printed photographs, replay attacks, and three-dimensional masks. Although recent deep learning-based face anti-spoofing (FAS) methods have achieved substantial improvements, many existing approaches still involve considerable computational cost, provide limited emphasis on fine-grained spoof artifacts, and face challenges in effectively integrating heterogeneous representations. To address these limitations, this paper proposes a Lightweight Face Anti-Spoofing Framework with Multi-Modal Representation, Spoof Artifact Enhancement, and Adaptive Feature Fusion. The framework starts from a single RGB facial image and constructs complementary Depth and Near-Infrared (NIR) representations using a Depth and Near-Infrared Construction Module (DNCM), rather than requiring dedicated Depth or NIR sensors. A shared EfficientNetV2 backbone is then employed to extract features from the three representations with reduced computational redundancy. The proposed Spoof Artifact Enhancement Module (SAEM) emphasizes subtle spoof-specific visual cues, while Cross-Modal Consistency Learning (CMCL) reduces representation discrepancies across the constructed modalities. Subsequently, the Adaptive Feature Fusion Module (AFFM) dynamically weights the refined representations according to their discriminative contribution. Extensive experiments on CelebA-Spoof, CASIA-SURF, HQ-WMCA, and SiW-M demonstrate the effectiveness of the proposed framework. The framework achieves accuracies of 98.16%, 97.54%, 99.08%, and 97.18%, with corresponding ACER values of 2.87%, 4.36%, 2.03%, and 4.18%, respectively. Across five independent runs, statistical analysis further indicates consistent performance with significant improvements over the selected baseline. In addition, the framework requires only 9.8 million parameters and 2.1 GFLOPs and achieves an inference speed of 82 FPS, demonstrating a favourable balance between detection effectiveness and computational efficiency. The results indicate that multi-modal representation combined with explicit spoof artifact enhancement and adaptive feature fusion can provide an efficient solution for robust and real-time face anti-spoofing.
Free
Scientific article
Efficient audio compression with low computational complexity is essential for real-time embedded systems operating under tough latency, memory, and hardware resource constraints. This paper presents a low-complexity hybrid audio compression framework that integrates Linear Predictive Coding (LPC), discrete wavelet transform (DWT) and Golomb entropy coding into a unified pipeline-oriented architecture for real-time FPGA implementation. The proposed framework uses LPC for short-term spectral modeling and residual extraction, Daubechies 4 wavelet transform for multi-resolution energy compaction and Golomb entropy coding for efficient compression of the resulting coefficients. The entropy coding scheme is lightweight and agreeable to hardware implementation. The architecture uses fixed-point arithmetic and pipelined processing to provide deterministic execution with low computational complexity on an Artix-7 FPGA platform. Experimental evaluation was performed on a 16 kHz uncompressed speech signal with 20 ms frames. The proposed framework achieved a compression ratio of 5.51 which is higher than that of LPC only (2.00), wavelet only (2.00) and MP3 (5.33) under the same evaluation conditions. The hardware implementation only used 3.83% LUT utilization, 0.94% flip-flop utilization and one DSP block. The processing latency of 0.037 ms per frame is significantly less than the 20 ms frame duration for real-time operation. Objective evaluation yielded SNR of 69.37 dB, STOI of 0.990, and PESQ of 2.19. This demonstrates that the suggested framework favors compression efficiency and hardware simplicity at the cost of reasonable reconstruction quality. The proposed LPC-Wavelet-Golomb architecture provides a practical compromise between compression performance, implementation complexity and real-time FPGA suitability for embedded audio compression applications.
Free
Scientific article
The research is concerned with the development of a mathematical model for estimating Intelligent Quotient of human above 65years. The model was optimized to know whether we have maximum or minimum human IQ level. However, the optimization result showed a saddled point which indicates that there is no minimum or maximum human IQ in life. This result implies that there is no particular IQ level that any human above 65years cannot attain when there is an enabling environment for adult education and other adult related trainings geared towards sharpening IQ skills of our retired and aged people. Similarly the IQ model was also validated with data from real life. And the outcome of the validation gave a correlation coefficient of 0.990405. This implies that our model is approximately 99% in agreement with real life data used, thus shows that the model is a standard measure for estimating IQ of human above 65years.
Free
A Measurable Approach for Access and Control Modeling in SOA
Scientific article
Service oriented architects (SOA) is considered as an extensible, robust and platform independent web application architect. However, problems in security guard for service access remains unresolved especially for the measurable one. we proposed a novel access control model, which we called SACM: Service Access Control Model, specially for SOA. Our model is mainly based on the role access policy, extended with trust authority transition and integration mechanism, to fulfill an extensive and measurable access control modeling approach with Crypto-CCS.
Free
A Methodical Study for Time-Frequency Analysis Model with Experimental Case Study on Chirp Signal
Scientific article
In this paper, we are reporting on the comprehensive model design for time-frequency analysis system using Short-Time Fourier Transform (STFT) and Wigner-Ville Distribution (WVD) methods. As a case study, both STFT and WVD based time-frequency transforms have been developed via MATLAB platform and applied for both Chirp and Sunspot signals. The developed model considers the use of hamming moving window of length L=50 with 90% overlapping between the current and previous window positions. The simulation results showed that WVD is more accurate method for time and frequency analysis than STFT since it can provide simultaneous localization in both time and frequency with higher resolution than STFT which can only provide localization in either time or frequency at the same time. Also, the applied techniques provide an adequate distribution of time-frequency analysis only if they used with a non-stationary signal such as Chirp signal.
Free
A Mobile-Integrated Deep Learning Framework for Early Detection of Maize Diseases
Scientific article
Maize is a cornerstone of food security and economic stability in Nigeria, yet its production is severely hampered by crop diseases that cause significant yield losses and threaten the livelihoods of millions of smallholder farmers. Despite advances in machine learning (ML) and deep learning (DL) for plant disease detection, existing solutions often lack generalizability, scalability, and accessibility for resource-limited settings. This research used a robust, predictive system that leverages convolutional neural networks, specifically ResNet50 and EfficientNet trained on diverse, annotated datasets of maize leaf images. By integrating computer vision, transfer learning, and user-centric mobile application design, the system aimed to provide real-time, accurate diagnosis and actionable recommendations for disease management. This study compared the performance of the ResNet50 and the EfficientNet. At the end of the research, ResNet50 achieved marginally higher accuracy than EfficientNet under the same experimental conditions, although the performance difference is small and not statistically tested. The ResNet50 model was thereafter deployed into a scalable mobile application tool that can empower farmers and extension workers with early disease detection capabilities, potentially reducing crop losses, improving productivity, and enhancing food security across sub-Saharan Africa.
Free
