Algorithm-based Methods Research Articles

Surface defect inspection of industrial products with object detection deep networks: a systematic review

One of the focal points in industrial product defect detection lies in the utilization of deep learning-based object detection algorithms. With the continuous introduction of these algorithms and their refined models, notable achievements have been attained. However, challenges persist in industrial settings, such as substantial variations in defect scales, the delicate balance between accuracy and speed, and the detection of small objects. Various methods have been proposed to address these challenges and propel the advancement of defect detection. To comprehensively review the latest developments in deep learning-based industrial product defect detection algorithms and foster further progress, this paper encompasses typical datasets and evaluation metrics used in industrial product defect detection, traces the development history of supervised one-stage and two-stage object detection algorithm-based and unsupervised algorithm-based industrial defect detection methods, discusses major challenges, and outlines future directions. It highlights the potential for further improving the accuracy, speed, and reliability of defect detection systems in industrial applications.

Multi-population differential evolution approach for feature selection with mutual information ranking

Feature selection is a crucial aspect of data preprocessing because of the significant effect of redundant features on classification performance and the extensive computational resources required. Evolutionary algorithm-based feature-selection methods have shown remarkable results in determining the optimal feature subset. To further enhance classification performance, this paper proposes a novel multi-population differential evolution approach for feature selection with mutual information ranking (MI-MPODE). Firstly, population preprocessing guided by mutual information is employed to reduce the dimensionality of the initial feature space. Then, the feature subset obtained from mutual information serves as the initial population for MI-MPODE. MI-MPODE incorporates a novel multi-population information-sharing mechanism, with common individuals from three layers contributing to inter-subpopulation information sharing. Additionally, an individual enhancement strategy is proposed to handle the variations of individuals in the population, and a Lens imaging opposition-based learning method is adopted to improve the algorithm’s optimization capability. MI-MPODE is compared with several state-of-the-art evolutionary algorithm-based feature selection methods through experimental comparisons. The results show that MI-MPODE outperforms all comparison algorithms on more than half of the datasets, with a significant reduction in the number of features used, demonstrating a significant advantage over competitors.

Deep Cascade of Convolutional Neural Networks for Quantification of Enlarged Perivascular Spaces in the Basal Ganglia in Magnetic Resonance Imaging.

In this paper, we present a cascaded deep convolution neural network (CNN) for assessing enlarged perivascular space (ePVS) within the basal ganglia region using T2-weighted MRI. Enlarged perivascular spaces (ePVSs) are potential biomarkers for various neurodegenerative disorders, including dementia and Parkinson's disease. Accurate assessment of ePVS is crucial for early diagnosis and monitoring disease progression. Our approach first utilizes an ePVS enhancement CNN to improve ePVS visibility and then employs a quantification CNN to predict the number of ePVSs. The ePVS enhancement CNN selectively enhances the ePVS areas without the need for additional heuristic parameters, achieving a higher contrast-to-noise ratio (CNR) of 113.77 compared to Tophat, Clahe, and Laplacian-based enhancement algorithms. The subsequent ePVS quantification CNN was trained and validated using fourfold cross-validation on a dataset of 76 participants. The quantification CNN attained 88% accuracy at the image level and 94% accuracy at the subject level. These results demonstrate significant improvements over traditional algorithm-based methods, highlighting the robustness and reliability of our deep learning approach. The proposed cascaded deep CNN model not only enhances the visibility of ePVS but also provides accurate quantification, making it a promising tool for evaluating neurodegenerative disorders. This method offers a novel and significant advancement in the non-invasive assessment of ePVS, potentially aiding in early diagnosis and targeted treatment strategies.

A Study on Countermeasures against Neutralizing Technology: Encoding Algorithm-Based Ransomware Detection Methods Using Machine Learning

Ransomware, which emerged in 1989, has evolved to the present in numerous variants and new forms. For this reason, serious damage caused by ransomware has occurred not only within our country but around the world, and, according to the analysis of ransomware trends, ransomware poses an ongoing and significant threat, with major damage expected to continue to occur in the future. To address this problem, various approaches to detect ransomware have been explored, with a recent focus on file entropy estimation methods. These methods exploit the characteristic increase in file entropy that is caused by ransomware encryption. In response, a method was developed to neutralize entropy-based ransomware detection technology by manipulating entropy using encoding methods from the attacker’s perspective. Consequently, from the defender’s standpoint, countermeasures are essential to minimize the damage caused by ransomware. Therefore, this article proposes a methodology that utilizes diverse machine learning models such as K-Nearest Neighbors (KNN), logistic regression, decision tree, random forest, gradient boosting, support vector machine (SVM), and multi-layer perception (MLP) to detect files infected with ransomware. The experimental results demonstrate empirically that files infected with ransomware can be detected with approximately 98% accuracy, and the results of this research are expected to provide valuable information for developing countermeasures against various ransomware detection technologies.

Building Geometry Survey by Using Ultra-Wideband (UWB) Wireless Technology and Algorithm-Based BIM Modelling

The efficiency and degree of maturity of the architectural survey determine the construction process in many ways. From the economic and technical point of view, it is significant. It can be found in almost all activities and, depending on the technology used, it provides different qualities and quantities of initial data and information to prepare and support, among other things, the design, demolition, and construction work processes. The objective of the research and development project and this article is to investigate and demonstrate the application of a novel solution that differs from the conventional building survey tools, methods, and procedures of the information. It breaks with traditional methods by implementing wireless (using Ultra-Wideband (UWB) technology) spatial positioning and algorithm-based post-processing methods, as well as Building Information Modelling (BIM).

Extraction of Tooth Cusps based on DBSCAN Density Clustering and Neighborhood Search Algorithm.

A good tooth cusp extraction is helpful in evaluating the effect of cosmetic dental work in virtual tooth surgery. We propose a new tooth cusp extraction, which integrates the DBSCAN (density-based spatial clustering of applications with noise) clustering algorithm with the neighborhood search algorithm to extract tooth cusp from a three-dimensional cloud-point tooth model. This method used the point cloud height and curvature to screen out the dented point set. Then we employ the DBSCAN clustering algorithm to segment different feature regions of the tooth surface and generate the candidate point set. Finally, the candidate point set was accurately located at the tooth apex through the neighborhood search algorithm and the traversal search method of non-maximum suppression. The experimental results show that the proposed method is superior to the traditional watershed algorithm-based methods by calculating the recall rate and accuracy rate, and also has higher extraction speed and extraction precision than manual extraction methods.

An Unsupervised Band Selection Method via Contrastive Learning for Hyperspectral Images

Band selection (BS) is an efficacious approach to reduce hyperspectral information redundancy while preserving the physical meaning of hyperspectral images (HSIs). Recently, deep learning-based BS methods have received widespread interest due to their ability to model the nonlinear relationship between bands, with existing methods typically relying on generative algorithms. However, the process of generating images with pixel-level detail required by generative algorithm-based BS methods is computationally expensive. To alleviate this issue, we propose a contrastive learning-based unsupervised BS architecture, termed ContrastBS, in this article. With the help of contrastive learning, the proposed architecture avoids the costly generation step in pixel space by learning to distinguish data at the abstract semantic level of the feature space. Specifically, ContrastBS combines an attention mechanism with contrastive learning to extract the importance of each band. Furthermore, we design a novel loss function, which is able to constrain the symmetric loss while ensuring attention to the most valuable bands, for the contrastive learning-based BS network. Experimental results indicate that ContrastBS has excellent classification performance and competitive time cost compared to the comparison methods.

Determinants of public preferences on low-carbon energy sources: Evidence from the United Kingdom

We empirically derive the determinants of British public preferences for different low-carbon energy sources using machine learning algorithm-based variable selection methods (ridge, lasso, and elastic net regression models). We seek to understand the drivers of support for solar, wind, biomass, and nuclear energy, which are the largest low-carbon energy sources and together account for the majority of UK power generation. Explanatory variables examined include those related to demographics, knowledge, perceptions of climate change, and government policy. We carry out a comparative study by synthesising the results of our independent analyses for each energy source and find that the preferred energy sources vary with respondents’ views on anticipated climate change impacts. Those who believe that potential effects of climate change will be catastrophic tend to prefer renewable energy sources whereas those less concerned about climate change tend to prefer nuclear power. The public also prefers energy sources about which they are more familiar or knowledgeable. Energy transition policies should be adjusted to better consider the factors that drive public acceptance of low-carbon energy sources, including exploring ways to secure policy support that considers energy source-specific characteristics and expanding public awareness of climate change and individual technologies, particularly among certain demographics.

X-Processes: Process model discovery with the best balance among fitness, precision, simplicity, and generalization through a genetic algorithm

Although process model discovery has been largely investigated over the past two decades, existing process discovery methods are not yet considered fully satisfactory. A particular issue is the difficulty of discovering process models with a good balance among recall (or fitness), precision, generalization, and simplicity, notably for real-world event logs. This article revisits the genetic algorithms-based X-Processes process discovery method, discussing the results of using a fitness function calculated through a harmonic mean that considers the four main metrics of process model quality used in process mining — recall, precision, generalization and simplicity. Although genetic algorithms have been used to discover process models, such methods have limitations, as do other non-genetic algorithm-based methods. X-Processes was tested with two harmonic mean options — one based on recall and precision only and another based on the four quality metrics. X-Processes showed superior effectiveness when compared to state-of-the-art process discovery methods, including one also based on genetic algorithms. The experimental results obtained for both harmonic mean options, using 12 real-world event logs, show that the overall quality of all process models discovered by X-Processes is superior to that of the baseline methods. While its execution time is longer than the other compared process discovery methods, X-Processes emerges as a solution when the need for a higher quality process model outweighs the hunger for agility.

Simplification of 3D CAD Model in Voxel Form for Mechanical Parts Using Generative Adversarial Networks

Most three-dimensional (3D) computer-aided design (CAD) models of mechanical parts, created during the design stage, have high shape complexity. The shape complexity required of CAD models reduces according to the field of application. Therefore, it is necessary to simplify the shapes of 3D CAD models, depending on their applications. Traditional simplification methods recognize simplification target shape based on a pre-defined algorithm. Such algorithm-based methods have difficulty processing unusual partial shapes not considered in the CAD model. This paper proposes a method that uses a network based on a generative adversarial network (GAN) to simplify the 3D CAD models of mechanical parts. The proposed network recognizes and removes simplification target shapes included in the 3D CAD models of mechanical parts. A 3D CAD model dataset was constructed to train the 3D CAD model simplification network. 3D CAD models are represented in voxel form in the dataset. Next, the constructed training dataset was used to train the proposed network. Finally, a 3D voxel simplification experiment was performed to evaluate the performance of the trained network. The experiment results showed that the network had an average error rate of 3.38% for the total area of the mechanical part and an average error rate of 14.61% for the simplification target area.

Review of Transformer Health Index from the Perspective of Survivability and Condition Assessment

As a critical indicator for assessing the survivability and condition of transformers in a fleet, the transformer health index has attracted attention from both asset owners and international organizations like CIGRE and IEEE DEIS/PES. To provide a systematic and comprehensive review for further study or to guide transformer asset management, this paper summarizes the state-of-the-art of the transformer health index, from the early proposed weighted-score-sum approaches to the more recently proposed artificial intelligence algorithm-based methods. Firstly, different methods for determining the transformer health index are reviewed. Each of these is specified as belonging to a certain type on the basis of its formulation and composition schematic. Subsequently, the steps to determine each type of health index are summarized, and examples derived from literature are provided for further illustration. Comparisons are finally carried out in order to better understand the pros and cons of different types of transformer health index, and the future development trends for transformer health indexes are also discussed. This work can serve as a valuable reference for the survivability and condition assessment of transformers in the power industry.

Named entity recognition and dependency parsing for better concept extraction in summary obfuscation detection

Summary obfuscation is a type of idea plagiarism where a summary of a text document is inserted into another text document so that it is more difficult to detect with ordinary plagiarism detection methods. Various methods have been developed to overcome this problem, one of which is based on genetic algorithms. This paper proposes a new approach for summary obfuscation detection based on named entity recognition and dependency parsing, which is straightforward but accurate and easy to analyze compared to genetic algorithm-based methods. The proposed method successfully detects summary obfuscation at the document level more accurately than existing genetic algorithm-based methods. Our method produced accuracy at sentence level up to more than 84% for specific benchmark and threshold cases. In addition, we have also tested our proposed method on other types of plagiarism, and the resulting accuracy is excellent.

Improved FAST algorithm for non-uniform rotational distortion correction in OCT endoscopic imaging.

Optical Coherence Tomography (OCT) is widely used for endoscopic imaging in endoluminal organs because of its high imaging accuracy and resolution. However, OCT endoscopic imaging suffers from Non-Uniform Rotational Distortion (NURD), which can be caused by many factors, such as irregular motor rotation and changes in friction between the probe and the sheath. Correcting this distortion is essential to obtaining high-quality Optical Coherence Tomography Angiography (OCTA) images. There are two main approaches for correcting NURD: hardware-based methods and algorithm-based methods. Hardware-based methods can be costly, challenging to implement, and may not eliminate NURD. Algorithm-based methods, such as image registration, can be effective for correcting NURD but can also be prone to the problem of NURD propagation. To address this issue, we process frames by coarse and fine registration, respectively. The new reference frame is generated by filtering out the A-scan that may have the NURD problem by coarse registration. And the fine registration uses this frame to achieve the final NURD correction. In addition, we have improved the Features from Accelerated Segment Test (FAST) algorithm and put it into coarse and fine registration process. Four evaluation functions were used for the experimental results, including signal-to-noise ratio (SNR), peak signal-to-noise ratio (PSNR), mean squared error (MSE), and structural similarity index measure (SSIM). By comparing with Scale-invariant feature transform (SIFT), Speeded up robust features (SURF), Oriented FAST and Rotated BRIEF (ORB), intensity-based (Cross-correlation), and Optical Flow algorithms, our algorithm has a higher similarity between the corrected frames. Moreover, the noise in the OCTA data is better suppressed, and the vascular information is well preserved. Our image registration-based algorithm reduces the problem of NURD propagation between B-scan frames and improves the imaging quality of OCT endoscopic images.

6D object position estimation from 2D images: a literature review

The 6D pose estimation of an object from an image is a central problem in many domains of Computer Vision (CV) and researchers have struggled with this issue for several years. Traditional pose estimation methods (1) leveraged on geometrical approaches, exploiting manually annotated local features, or (2) relied on 2D object representations from different points of view and their comparisons with the original image. The two methods mentioned above are also known as Feature-based and Template-based, respectively. With the diffusion of Deep Learning (DL), new Learning-based strategies have been introduced to achieve the 6D pose estimation, improving traditional methods by involving Convolutional Neural Networks (CNN). This review analyzed techniques belonging to different research fields and classified them into three main categories: Template-based methods, Feature-based methods, and Learning-Based methods. In recent years, the research mainly focused on Learning-based methods, which allow the training of a neural network tailored for a specific task. For this reason, most of the analyzed methods belong to this category, and they have been in turn classified into three sub-categories: Bounding box prediction and Perspective-n-Point (PnP) algorithm-based methods, Classification-based methods, and Regression-based methods. This review aims to provide a general overview of the latest 6D pose recovery methods to underline the pros and cons and highlight the best-performing techniques for each group. The main goal is to supply the readers with helpful guidelines for the implementation of performing applications even under challenging circumstances such as auto-occlusions, symmetries, occlusions between multiple objects, and bad lighting conditions.

A Hybrid Multi-Objective Optimization Method and Its Application to Electromagnetic Device Designs

Optimization algorithms play a critical role in electromagnetic device designs due to the ever-increasing technological and economical competition. Although evolutionary algorithm-based methods have successfully been applied to different design problems, these methods exhibit deficiencies when solving complex problems with multimodal and discontinuous objective functions, which is quite common in electromagnetic device optimization designs. In this paper, a hybrid multi-objective optimization algorithm based on a non-dominated sorting genetic algorithm (NSGA-II) and a multi-objective particle swarm optimization method (MOPSO) is proposed. In order to enhance the convergence and diversity performance of the algorithm, a new population update mechanism of MOPSO is introduced. Moreover, an adaptive operator involving crossover and mutation is presented to achieve a better balance between global and local searches. The performance of the hybrid algorithm is validated using standard test functions and the multi-objective design of a superconducting magnetic energy storage (SMES) device. Numerical results demonstrate the effectiveness and superiority of the proposed method.

Sustainable cooling cycles by algorithmically supported design of decentral pump systems

Accounting for up to 25% of electricity consumption in industry, pump systems are significant electricity consumers. Designing pump systems is a complex process because of the multitude of parameters and their combinations (pump types, interconnection variants, various load scenarios). Moreover, decentral pumps can be integrated to reduce the required hydraulic power, which further increases the complexity. This complexity can no longer be mastered with conventional planning methods. The aim of this paper is to tackle the complexity of pump systems design using algorithm-based design methods. The design problem is formulated as a mixed-integer nonlinear optimisation program with the objective of minimising the life cycle costs. The solution involves three steps: (i) the placement of decentral pump stations to reduce the hydraulic power, (ii) the preselection of suitable pump types, and (iii) the optimisation of the selection and operation of pumps. To solve the optimisation problem, an exact off-the-shelf solver and a problem-specific heuristic solution method are used. These are compared with conventional design methods. Both the algorithm-based and the conventional design methods are applied to an industrial cooling cycle operated by BASF SE. The results show that due to decentralisation, up to 38% (248 kW) of hydraulic power can be saved. With the help of the algorithm-based design methods, savings can also be achieved with regard to the efficiency and costs: up to 21% of the life-cycle costs can be saved while increasing the net efficiency from 47.2% to 57.8%. Furthermore, the algorithm-based design methods are superior, leading to 15% (approx. 770 000 €) lower life-cycle costs compared to conventional methods. Conventional methods based on the best efficiency point or mean efficiency show the lowest level of performance. The advantages are particularly evident in system variants with a high complexity. The trade-off between investment costs and energy demand is presented transparently using a Pareto front.

An intelligent hybrid method: Multi-objective optimization for MEC-enabled devices of IoE

Emerging Internet-of-Everything (IoE) services require connected devices to respond instantly and operate for long durations. As smart mobile devices (SMDs) are often powered by batteries of limited capacity, offloading some computational tasks to nearby edge servers is a promising solution to reduce the latency and energy consumption of SMDs in operation. However, a challenge of computation offloading in the IoE system is the large number of SMDs that need to be handled. To address this problem, in this paper, we propose an intelligent two-stage computation offloading scheme with multiple optimization objectives. In the first stage, we categorize the computation tasks into classes (e.g., computation-intensive, data-intensive) and make early offloading decisions on some classes of tasks with offloading preferences. In the second stage, we make offloading decisions on the remaining tasks by solving a multi-objective optimization problem using the powerful Non-dominated Sorting Genetic Algorithm (NSGA-II). This two-stage design can help reduce the size of task instances in the second optimization stage, thus accelerating the convergence of the NSGA-II algorithm. The extensive simulation results show that, compared to the existing NSGA-II algorithm-based optimization methods, the proposed offloading scheme improves the performance by 10% in terms of latency and energy consumption.

Performance Improvement of Genetic Algorithm Based Exam Seating Solution by Parameter Optimization

Exam seat allocation becomes a complex problem concerning the increasing number of students, subjects, exams, departments, and rooms in a higher institution. The requirements and constraints of this problem demonstrate similar characteristics to extensively researched exam timetabling problems. They are planning limited capacity effectively and efficiently. Additionally, exam seating requires a seating arrangement to reduce cheating incidents. In the literature, several genetic algorithm-based methods were recommended to prevent students that have a friendship from sitting close during the exams while providing the best exam session arrangement. We have improved the performance of the genetic algorithm by parameter optimization and a new elitism method to increase saturation rate and accuracy. The algorithm has been tested on a real-world data set and represents a high potential for the realization of a high-quality seating arrangement compatible with the real world.

Cycle temporal algorithm-based multivariate statistical methods for fault diagnosis in chemical processes

Multivariate statistical process monitoring methods are often used in chemical process fault diagnosis. In this article, (I) the cycle temporal algorithm (CTA) combined with the dynamic kernel principal component analysis (DKPCA) and the multiway dynamic kernel principal component analysis (MDKPCA) fault detection algorithms are proposed, which are used for continuous and batch process fault detections, respectively. In addition, (II) a fault variable identification model based on reconstructed-based contribution (RBC) model that paves the way for determining the cause of the fault are proposed. The proposed fault diagnosis model was applied to Tennessee Eastman (TE) process and penicillin fermentation process for fault diagnosis. And compare with other fault diagnosis methods. The results show that the proposed method has better detection effects than other methods. Finally, the reconstruction-based contribution (RBC) model method is used to accurately locate the root cause of the fault and determine the fault path.

The Clinical Application of Remimazolam Benzenesulfonate Combined with Esketamine Intravenous Anesthesia in Endoscopic Retrograde Cholangiopancreatography.

In this project, algorithm-based image processing methods in 3D endoscopic image processing endoscopic retrograde cholangiopancreatography (ERCP) were analyzed. To enhance local information of images, an adaptive histogram equalization method with limited contrast is introduced. The influences of the algorithm on 3D endoscopic image peak signal-to-noise ratio (PSNR), image discrete information entropy (DE), and average mean brightness error (AMBE) of images before and after the optimization before were compared. A total of 92 patients receiving ERCP at Yuhuangding Hospital between December 2019 and December 2021 were selected and divided into the control group (fentanyl+propofol) and the observation group (remimazolam benzenesulfonate+esketamine). Mean arterial pressure heart rate (HR), oxygen saturation (SpO2), and respiratory rate (RR) of the patients at each time point including the entry into the operation room (T0), 2 minutes after the beginning of medication (T1), after endoscopy (T2), endoscopy withdrawal (T3), and postoperative awakening (T4) were recorded. The comparison of MAP between T1, T2, T3, and T4 and T0 among patients in the observation group and the control group showed statistical differences (P < 0.05). Besides, HR and RR at T4 in the observation group were obviously higher than those in the control group (P < 0.05). The comparison of SpO2 at T3 and T4 and that at T0 both showed statistical differences (P < 0.05). Awakening time and VAS scores in the observation group were obviously lower than those in the control group (P < 0.05). The incidence of bradycardia, nausea, vomiting, and chill in the observation group was all lower than that in the control group (P < 0.05). The results indicated that an effective endoscopic image processing method was established based on an image enhancement algorithm, and the combination of remimazolam benzenesulfonate and esketamine showed high safety and efficacy in ERCP.