International Journal of Engineering and Manufacturing @ijem
Journal articles - International Journal of Engineering and Manufacturing
All articles: 649
DataViz Model: A Novel Approach towards Big Data Analytics and Visualization
Scientific article
Big Data is the collection of large data sets which contains large amount of data. There are different areas which are generating huge data, this data may be present in the form of semi-structured or unstructured and to get useful information from such raw data there is need of data analysis. Due to Big Data’s excessive volume, variety, and velocity it is very difficult to store and process huge data. The process of extracting the information from such raw data is called Big Data Analytics. Big data Analytics processes data gives result in the form of structured data. Again this data is huge size and very difficult to understand since it is present in the form of CSV or excel or simple text files. So for effective decision making and to understand the information quickly the data need to be visualized as human mind understands images and graphs better and faster than text data. In this paper a model called Data Visualization (Viz) is designed which integrates big data analytics and the data visualization. This model first takes the data from various sources and then processes it and converts it into structured form, if want this data can be stored to RDBMS. Finally the text result can be visualized with the help of Visualization module of the DataViz. Here text result is represented in the form of charts and graphs.
Free
Database Design of a General Data Analysis System of Commodity Sales Information
Scientific article
A data analysis system for the general commodity sales information is researched on and further designed. The key problems of designing this system are that, it should enable the system to adjust to user-data with different structures, enable users to define or change data structures as well as retrieval methods. Through the research on the general structure and retrieval method of the commodity sales data, the system realizes users' customization of the database, thus is applicable to sales data with different structures. This paper focuses on the structure of the database, the creation method and process of the database by the user, the structure of the data dictionary and data exchange between database and software, with case examples in the final illustration.
Free
DbDMAD: a Database of DNA Methylation in Human Age-related Disease
Scientific article
DNA methylation plays a variety of crucial roles in cell division, proliferation, development of aging life, development of genetic diseases related to uniparental disomy, and carcinogenesis. DNA methylation can be probed by HPLC and gene chip which is very helpful to the high-through methylation test. In recent years many published articles reported that DNA methylation may be linked with human aged disease. Mining and integration of DNA methylation in human aged disease can be beneficial to novel biological discoveries. There has not been DNA methylation database repository which is exclusive for human aged disease. Therefore, we developed dbDMAD: a database of DNA methylation in human aged diseases, there are two purposes of it, one is to store DNA methylation in human aged disease datasets which were obtained from laboratory experiments, another is to find the relationship of different aged diseases. This is the first release of dbDMAD in which users can find 12 kinds of human aged diseases and relevant DNA methylation information. It can be searched by disease name and gene ID. This database also includes a visualization tool named ChainMap, by which the map of methylation pathway can be shown.
Free
Deep Convolution Neural Networks for Cross-Dataset Facial Expression Recognition System
Scientific article
Facial Expressions are a true and obvious way to represent emotions in human beings. Understanding facial expression recognition (FER) is essential, and it is also useful in the area of Artificial Intelligence, Computing, Medical, Video games, e-Education, and many more. In the past, much research was conducted in the domain of FER using different approaches such as analysis through different sensor data, using machine learning and deep learning framework with static images and dynamic sequence. Researchers used machine learning-based techniques such as the Multi-layer Perceptron Model, k- Nearest Neighbors, and Support Vector Machines were used by researchers in solving the FER. These methods have extracted features such as Local Binary Patterns, Eigenfaces, Face-landmark features, and Texture features. Recently use of deep learning algorithms in FER has been considerable. State-of-the-art results show deep learning-based approaches are more potent than conventional FER approaches. This paper focuses on implementing three different Custom CNN Architecture training them on FER13 Dataset and testing them on CK+ and JAFFE Dataset including FER13 after fine-tuning. The three pre-trained models' on FER2013 after fine-tuning have significantly improved the accuracy of the resulting CNN on the target test sets between 65.12 % to 79.07% on the JAFFE dataset and 50.96% to 68.81% on the CK+ dataset.
Free
Scientific article
Fashion image analysis has a lot of trouble working with noisy data, especially when there aren't any good training examples to use. This happens because standard auto encoders can't keep class separability when the noise level changes. This paper introduces Structured Auto encoders (SAE), which combine convolutional encoder-decoder architectures with geometric constraints in the latent space. This strategy uses greedy layer-wise pre-training, and then it fine-tunes the model using two loss functions: structured latent space regularization and reconstruction error. The encoder has convolutional layers with 3×3 filters and ReLU functions that only turn on when they are needed. During testing, it was tested on Fashion-MNIST with noise levels ranging from 20% to 70%. It was also tested on MNIST, DeepFashion2, and 3D human pose datasets. The results of the experiment show that SAE can classify 85.4% of the time with only five samples labeled for each group. This is much better than the standard auto encoders (68.4%) and the baseline support vector machine (SVM) methods (84.32%). With the best setup, which has 1,000 hidden nodes and a learning rate of 0.1, images with up to 70% noise can be reconstructed well With 80% classification accuracy on clean test data, the latent space structured method allows for the solution of basic issues that are related to introducing geometric constraints between the classes so that the class boundaries remain intact even in a noisy environment.The proposed method achieves a classification accuracy of 85.4% ± 1.2%under standard evaluation settings, with robustness validated under noise levels of up to 70%. All results are obtained using standard training–testing splits and are averaged over multiple experimental runs to ensure reliability and reproducibility. The proposed semi-supervised performance is enhanced by increasing the number of discriminative features through the use of a highly structured representation of the data, as opposed to using an unstructured representation, which results in performance enhancement. The author presents an original framework capable of simultaneously providing robust image classification and denoising capability through geometric constraints established in the framework, which enhances the learning of discriminative features through the use of limited amounts of labeled data. The implementation of this framework in the real-world fashion industry would facilitate the processing of fashion images with respect to noise resistance and the ability to annotate images quickly.
Free
Deep Hybrid Neural Network for Automated Epileptic Seizure Detection from EEG Signals using MAT-LAB
Scientific article
Epilepsy is a long-term neurological disorder marked by recurring seizures resulting from irregular neuronal activity in the brain. Prompt and precise identification of epileptic events from electroencephalogram (EEG) signals is essential for successful clinical diagnosis. This study introduces a Deep Hybrid Neural Network framework that combines Convolutional Neural Networks (CNN) with the Aquila Optimizer (AO) for the automatic detection of epileptic seizures utilizing EEG data in MATLAB. The suggested system initially converts EEG signals into time–frequency spectrograms through Short-Time Fourier Transform (STFT), allowing the CNN to capture advanced spatial–spectral characteristics. The AO algorithm further improves these features by tuning hyperparameters and choosing the most distinguished fea-ture subsets, thereby boosting classification accuracy and decreasing computational overhead. The Bonn University EEG dataset was used to evaluate the model through a 5-fold cross-validation method, attaining an average accuracy of 95.62%, where per-class sensitivity and specificity surpassed 97%. Comparative evaluation showed that the CNN–AO hybrid sur-passed traditional classifiers in terms of accuracy and convergence reliability. These findings demonstrate the effectiveness of the proposed hybrid framework for automated epileptic seizure detection and suggest its potential suitability for future real-time and wearable healthcare applications following further deployment-oriented validation.
Free
Deep Learning Method for Early Detection of Lung Cancer using Deep Learning Algorithms
Scientific article
Lung cancer is responsible for many deaths from cancer around the globe, primarily because it is difficult to find malignant lung nodules early enough to be treatable. We developed a hybrid deep learning approach to the automated classification of lung nodules from chest computed tomography (CT) images. Our model uses convolutional neural networks (CNNs) for hierarchical feature extraction, an attention mechanism for feature refinement in targeted regions of interest, and a support vector machine (SVM) classifier for robust margin-based decision making. Furthermore, we use a patch-based learning strategy within the model to improve sensitivity to small and ambiguous lung nodules. The model is tested on the publicly available LIDC-IDRI dataset and achieves 94.2% accuracy, 95.1% recall, and an area under the receiver operating characteristic curve (AUC-ROC) score of 0.971, which outperforms multiple baseline deep learning methods. The proposed method provides a synergistic integration of attention-weighted feature enhancements and traditional machine learning classifications as compared to traditional end-to-end architectures, resulting in improved model generalization and interpretability. Grad-CAM visualizations are also used to provide qualitative insights into the model decision-making process. The proposed hybrid approach provides a novel and interpretable solution for the classification of lung nodules from CT images that may assist in the development of computerized systems to assist physicians in making diagnoses using medical images.
Free
Deep Learning Models for Non-Invasive Blood Group Detection Using Fingerprint Images
Scientific article
In the context of medical diagnosis, the identification of human blood groups plays a significant role. To perform the identification of human blood groups, usually invasive identification methods are used. However, due to the limitations of the invasive methods of blood group identification, the use of fingerprint-based identification of human blood groups gained significance recently. In this paper, an accurate comparison of the recently developed deep learning models of fingerprint-based blood group identification techniques is provided. Five CNN-based models, such as ConvNeXt-Tiny, ConvNeXt-Small, EfficientNetV2-S, RepVGG-B3g4, and MobileViT-V2 models for the identification of human blood groups, are implemented. The results obtained in the experiment, considering the accuracy of the models, have proved the EfficientNetV2-S model to have the highest accuracy of 98.84%, followed by the ConvNeXt-Small, ConvNeXt-Tiny, MobileViT-V2, and RepVGG-B3g4 models with accuracies of 97.66%, 97.21%, 97.08%, and 96.98%, respectively.
Free
Deep Learning based Real Time Radio Signal Modulation Classification and Visualization
Scientific article
Radio Modulation Classification is implemented by using the Deep Learning Techniques. The raw radio signals where as inputs and can automatically learn radio features and classification accuracy. The LSTM (Long short-term memory) based classifiers and CNN (Convolutional Neural Network) based classifiers were proposed in this paper. In the proposed work, two CNN based classifiers are implemented such as the LeNet classifier and the ResNet classifier. For visualizing the radio modulation, a class activation vector (w) is used. Finally in the proposed work, it is performed the classification by using the Deep learning models like CNN and LSTM based modulation classifiers. These deep learning models extract the important radio features that are used for classification. Here, the bench mark dataset RadioML2016.10a is used. This is an open dataset which contains the modulated signal I and Q values fewer than ten modulation categories. After evolution of proposed model with bench mark dataset, it is applied with real time data collected through the SDR Dongle receiver. From the obtained real time signal, the modulation categories have been classified and visualized the radio features extracted from the radio modulation classifiers.
Free
Deep Learning-Based Potato Leaf Disease Detection Using CNN in the Agricultural System
Scientific article
Potatoes play a vital role as a staple crop worldwide, making a significant contribution to global food security. However, the susceptibility of potato plants to various leaf diseases poses a threat to crop yield and quality. Detecting these diseases accurately and at an early stage is crucial for the effective management and protection of crops. Recent advancements in Convolutional Neural Networks (CNNs) have demonstrated potential in image categorization applications. Therefore, the goal of this work is to investigate the potential of CNNs in detecting potato leaf diseases. As neural networks have become part of agriculture, numerous researchers have worked on improving the early detection of potato blight using different machine and deep learning methods. However, there are persistent problems related to accuracy and the time it takes for these methods to work. In response to these challenges, we tailored a convolutional neural network (CNN) to enhance accuracy while reducing the trainable parameters, computational time and information loss. To conduct this research, we compiled a diverse dataset consisting of images of potato leaves. The dataset encompassed both healthy leaves and leaves infected with common diseases such as late blight and early blight. We took great care in curating and preprocessing the dataset to ensure its quality and consistency. Our focus was to develop a specialized CNN architecture tailored specifically for disease detection. To improve the performance of the network, we employed techniques like data augmentation and transfer learning during the training phase. The experimental outcomes demonstrate the efficacy of our proposed customized CNN model in accurately identifying and classifying potato leaf diseases. Our model's overall accuracy was an astounding 99.22%, surpassing the performance of existing methods by a significant margin. Furthermore, we evaluated precision, recall, and F1-score to evaluate the model's effectiveness on individual disease classes. To give an additional understanding of the model's behavior and its capacity to distinguish between various disease types, we utilized visualization techniques such as confusion matrices and sample output images. The results of this study have implications for managing potato diseases by offering an automated and reliable solution for early detection and diagnosis. Future research directions may include expanding the dataset, exploring different CNN architectures, and investigating the generalizability of the model across different potato varieties and growing conditions.
Free
Deep Learning-Based Pothole Detection Techniques under Multiple Weather Conditions
Scientific article
Potholes are a major concern for road infrastructure, traffic safety, and vehicle maintenance. Manual inspection methods for pothole detection are labor-intensive, time-consuming, and often inefficient for large road networks. This study evaluates and compares the performance of YOLOv5 and Single Shot Detector (SSD) models for automated pothole detection under diverse weather and lighting conditions. Using the Multi-Weather-Based Dataset (MWBD), images captured during daytime, twilight, and nighttime were annotated with bounding boxes and enhanced through data augmentation techniques such as shearing and flipping. Experimental results indicate that YOLOv5 achieves a precision of 92.2%, recall of 89.2%, F1-score of 90.7%, and mAP@0.5 of 90.0%, while SSD achieves a precision of 88.5%, recall of 92.0%, F1-score of 90.2%, and mAP@0.5 of 91.4%. The comparative analysis demonstrates that both models are effective in detecting potholes across varied road textures and environmental conditions, with trade-offs between precision and recall. This study highlights the suitability of deep learning-based object detection models for automated road inspection, reducing human effort, enhancing maintenance efficiency, and improving road safety. The novelty lies in the systematic comparison of YOLOv5 and SSD under multi-weather conditions, providing practical guidance for intelligent transportation systems.
Free
Scientific article
Handwritten character recognition is a crucial challenge in artificial intelligence and computer vision, particularly for complex scripts like Devanagari. Devanagari is widely used in Hindi, Marathi, and Nepali and consists of intricate multilevel compound characters, ligatures, and highly variable handwriting styles. Despite advances in optical character recognition (OCR) technology, accurately recognizing handwritten Devanagari characters remains difficult. This study compares various deep learning models, including Convolutional Neural Networks (CNN), CNN-SVM, Long Short-Term Memory (LSTM), EfficientNet, and a newly proposed attention-based CNN model. Extensive experiments were conducted on a diverse dataset containing simple and compound handwritten Devanagari characters. The proposed attention-based CNN model outperforms traditional methods, achieving a recognition accuracy of 96.5%, significantly higher than CNN (88.0%), CNN-SVM (88.5%), LSTM (92.0%), and EfficientNet (93.0%). The study employs advanced data augmentation techniques to enhance model robustness, making it adaptable to various handwriting styles. The attention mechanism in the proposed CNN model allows for improved feature extraction, leading to higher recognition accuracy and efficiency. This research contributes to developing robust OCR systems for the Devanagari script, enabling improved digitisation and preservation of Indian languages. The proposed approach can be extended to other complex scripts like Bengali and Tamil, further advancing multilingual OCR technologies, is now exploratory and has not been experimentally verified in this study. Future work can address the need for thorough cross-script evaluation and transfer learning studies to verify the adaptability of the attention-based CNN architecture, despite its inherent script-agnostic nature. The findings of this study hold significant implications for text digitization, historical document preservation, and automated language processing applications.
Free
Deep neural network for human face recognition
Scientific article
Face recognition (FR), the process of identifying people through facial images, has numerous practical applications in the area of biometrics, information security, access control, law enforcement, smart cards and surveillance system. Convolutional Neural Networks (CovNets), a type of deep networks has been proved to be successful for FR. For real-time systems, some preprocessing steps like sampling needs to be done before using to CovNets. But then also complete images (all the pixel values) are passed as input to CovNets and all the steps (feature selection, feature extraction, training) are performed by the network. This is the reason that implementing CovNets are sometimes complex and time consuming. CovNets are at the nascent stage and the accuracies obtained are very high, so they have a long way to go. The paper proposes a new way of using a deep neural network (another type of deep network) for face recognition. In this approach, instead of providing raw pixel values as input, only the extracted facial features are provided. This lowers the complexity of while providing the accuracy of 97.05% on Yale faces dataset.
Free
Scientific article
Refer to this research, an intelligent fuzzy parallel switching Proportional-Derivative (PD) plus gravity controller is proposed for highly nonlinear continuum robot manipulator. Design a nonlinear controller for second order nonlinear uncertain dynamical systems is one of the most important challenging works. In order to provide high performance in nonlinear systems, switching partly sliding mode plus gravity controller is selected. Pure switching partly sliding mode plus gravity controller can be used to control of partly known nonlinear dynamic parameters of continuum robot manipulator. Conversely, this method is used in many applications; it must to solve chattering phenomenon which it can cause some problems such as saturation and heat the mechanical parts of continuum robot manipulators or drivers. In order to solve the chattering phenomenon, implement easily and avoid mathematical model base controller, Mamdani’s performance/error-based fuzzy logic methodology with two inputs and one output and 49 rules is parallel applied to pure switching partly sliding mode plus gravity controller. The results demonstrate that this method is a model-free controllers which works well in certain and partly uncertain system.
Free
Scientific article
Refer to this research, an intelligent robust fuzzy parallel feedback linearization estimator for Proportional-Integral-Derivative (PID) controller is proposed for highly nonlinear continuum robot manipulator. In the absence of robot knowledge, PID may be the best controller, because it is model-free, and its parameters can be adjusted easily and separately. And it is the most used in robot manipulators. In order to remove steady-state error caused by uncertainties and noise, the integrator gain has to be increased. This leads to worse transient performance, even destroys the stability. The integrator in a PID controller also reduces the bandwidth of the closed-loop system. Model-based compensation for PD control is an alternative method to substitute PID control. Feedback linearization compensation is one of the nonlinear compensator. The first problem of the pure feedback linearization compensator (FLC) was equivalent problem in certain and uncertain systems. The nonlinear equivalent dynamic problem in uncertain system is solved by using parallel fuzzy logic theory. To eliminate the continuum robot manipulator system’s dynamic; Mamdani fuzzy inference system is design and applied to FLC. This methodology is based on design parallel fuzzy inference system and applied to equivalent nonlinear dynamic part of FLC. The results demonstrate that the model free fuzzy FLC estimator works well to compensate linear PID controller in presence of partly uncertainty system (e.g., continuum robot).
Free
Design Database for an Automated Control System of Typical Wastewater Treatment Processes
Scientific article
Current requests, that are advanced to the cleaning parameters of wastewaters, including the offered by the EU community standards, are quite high. To satisfy them, a lot of parameters need to be maintained at a proper level. The presence of a significant number of control points, and a wastewater treatment process management, many situations affect the cleaning process in the case of deviations from the specified conditions, as well as many pollutants indicate that currently structured information for quality cleaning processes management should be available. Even an experienced operator cannot handle the monitoring and management of so many parameters optimally without the relevant database. The modern phase of the water-saving technology enterprises is characterized by large volumes and complexity of the information being processed, by the need for public access to data elements, by high demands to the speed of the information acquisition and its reliability that can be realized under the conditions of the functioning database. Since the database should not only be a storage, but should also have a link between the different elements, the data structure must be described formally. For that reason, the database subschema was designed (as the conceptual data model) for automation of water-saving processes. With the help of normalization, the hierarchical database model has been reduced to relational. It has significant advantages over the hierarchical and network data structures including the following: clarity of data presentation, ability to use one relational algebra (a language of Codd algebra), as well as simplicity of database reorganization. Due to the lack of a recognized analytical method of database management system selection, the choice of database for automatic water-saving control processes was carried out based on an expert estimations method. To maintain the functioning of the database on the defined criteria, the PostgreSQL database management system has been selected. As a result, that will implement the interface with application software concerning the automation of wastewater treatment technological processes, as well as the main production automation.
Free
Scientific article
Preliminary studies conducted in the field of ultrasonic broadcast for bird pest control has identified among others, that ultrasound of specific frequencies and certain audio sounds can serves as effective parameters in weaver bird pest scaring. Of interest is ultrasound frequency of 35 kHz as the put-to-flight and stay-away frequency in bird homes, while 25 kHz holds in farm conditions. These findings set the stage for this work which is aimed at determining the compositions, dimensions and specifications of a stand-alone device capable of electronically generating ultrasound of the required frequencies and incorporating audio sounds of predators to effectively scare weaver birds away from farms while overcoming the challenge of habituation. The design consists of sections as: the solar power supply, tripping, timing, oscillatory, frequency selection, amplification, microcontroller, predator cry, LCD and ultrasonic transducer sections and their respective circuits. The design if implemented and tested, will ensure 360o horizontal ultrasound coverage and a vertical boost to effectively repel weave bird pests from the vicinity of broadcast.
Free
Design Modification and Fabrication of An Active Solar Dryer
Scientific article
This study presents the design modification and performance evaluation of an active solar dryer. The Modified Active Solar Dryer (MASD) is equipped with a non-concentrating flat plate thermal collector and drying chamber to dry various agricultural products. The drying chamber contains trays inside it. These are used to completely circulate airflow inside it. The drying Chamber has a cross-section Area of 0.30m × 0.45m and a height of 0.45m. The Flat plate collector has 0.91 m length, 0.45 m width, 0.12 m depth, and 0.049 m^3 volume. This study performed experimental work and evaluated performance by drying the chili in a Simplified Active Solar Dryer and Modified Active Solar dryer from 8 am to5 pm with average solar irradiance of 883.25W⁄m^2 . The temperature has been recorded at six different points. The difference in temperatures between the dryer input and the surrounding air has been observed continuously. Temperatures at the Thermal Collector and Drying Chamber are the main prime mover of solar drying, so our main concern was on these two temperatures. The temperature achieved by SASD in the thermal collector is in the range of 55 ~ 60˚C meanwhile in the MASD temperature range is 65 ~ 70˚C. In SASD chili was dried in five days meanwhile in Modified Active Solar Dryer it took four days. The moisture content of the chili was reduced to 10% and it took 45hour for SASD and 34 hours for MASD. It is concluded that Modified Active Solar Dryer took 34% less time than the Simplified Active Solar Dryer.
Free
Design Of A Optimized Parallel Array Multiplier Using Parallel Prefix Adder
Scientific article
Multiplication is the basic building block for several DSP processors, Image processing and many other. Over the years the computational complexities of algorithms used in Digital Signal Processors (DSPs) have gradually increased. This requires a parallel array multiplier to achieve high execution speed or to meet the performance demands. A typical implementation of such an array multiplier is Braun design. Braun multiplier is a type of parallel array multiplier. The architecture of Braun multiplier mainly consists of some Carry Save Adders, array of AND gates and one Ripple Carry Adder. In this research work, a new design of Braun Multiplier is proposed and this proposed design of multiplier uses a very fast parallel prefix adder (Brent kung Adder) in place of Ripple Carry Adder. The architecture of standard Braun Multiplier is modified in this work for reducing the area and delay due to Ripple Carry Adder and performing faster multiplication of two binary numbers. The design is implemented using Microwind1, digital schematics (DSCH)
Free
Design Remote Monitoring System for Patients at Real-Time based on Internet of Things (IoT)
Scientific article
The remote real-time patient monitoring system is a healthcare solution that uses ESP32 microcontroller and Blynk IoT cloud platform to monitor the vital signs of patients, including temperature, oxygen saturation, and heartbeat. The system also monitors the environmental factors surrounding the patient, such as temperature and humidity, and determines the GPS location of the patient. Additionally, the system includes an alarm device that alerts healthcare providers in case of emergency. In this paper we design system aims to provide continuous care and monitoring for patients, whether they are in hospitals, at home, or outside. By using Blynk IoT cloud platform, the system aims to reduce the percentage of medical errors and deaths by providing real-time monitoring of the patient's vital signs and environmental conditions, allowing healthcare providers to respond to emergencies quickly and efficiently. The IoT-based patient monitoring system consists of sensors that collect data on the patient's vital signs and environmental factors. The collected data is transmitted wirelessly to the Blynk IoT cloud platform, where it is processed and analyzed. Healthcare providers can access the data through the Blynk mobile app and receive alerts in case of any abnormalities or emergencies.
Free
