A significant amount of road maintenance cost goes into pothole repairs. The primary cost factors related to potholes are their size and depth, as larger and thicker potholes incur higher repair costs. However, existing methods for estimating pothole repair in developing countries rely on manual size measurements, which is time consuming, labor intensive, subjective and can lead to poor estimation of repair cost. This paper presents a system that can automatically determine the size of potholes from digital images and estimate the cost of repair. In this study, the stereo vision method was used to automatically estimate the depths of potholes from digital camera images. A feed-forward backward propagation Artificial Neural Network (ANN) was trained using pothole images acquired using mobile phones. The predicted depths and sizes of the potholes were then used to estimate the quantity of materials required to fill the potholes and subsequently, the cumulative cost of repair. Marking out and manual size measurements were performed for twenty randomly selected potholes in the Ugbowo Campus of the University of Benin, Nigeria. These measurements were compared against the estimated sizes of potholes predicted by the ANN model. A system was developed to automatically compute these material costs and considering other cost components such as transportation, labor, and equipment. Results obtained showed that the mean errors for depth, width and height estimates were 3.403%, 3.789% and 5.2617% respectively. Consequently, the developed system correctly estimated the cost of repair of the potholes considered in this study. A significant contribution of the paper is the speed and convenience of acquiring pothole data using a mobile phones without the need for on spot assessment of potholes or use of relatively more expensive stereoscopic camera setup.
大量的道路维护费用用于坑洞维修。与坑洞有关的主要成本因素是其大小和深度，因为较大和较厚的坑洞会产生较高的维修成本。然而，发展中国家现有的坑洞维修估算方法依赖于人工尺寸测量，这种方法耗时、耗力、主观，而且可能导致维修成本估算不准确。本文介绍了一种可从数字图像中自动确定坑洞大小并估算修复成本的系统。本研究采用立体视觉方法从数码相机图像中自动估算坑洞的深度。使用手机获取的坑洞图像训练了一个前馈后向传播人工神经网络（ANN）。然后，利用预测的坑洞深度和大小来估算填补坑洞所需的材料数量，进而估算出累计维修成本。对尼日利亚贝宁大学 Ugbowo 校区随机选取的 20 个坑洞进行了标记和人工尺寸测量。这些测量结果与 ANN 模型预测的坑洞估计尺寸进行了比较。结果显示，深度、宽度和高度估计值的平均误差分别为 3.403%、3.789% 和 5.2617%。因此，所开发的系统正确估算了本研究中考虑的坑洞修复成本。本文的一个重要贡献是，使用手机获取坑洞数据既快捷又方便，无需对坑洞进行现场评估，也无需使用相对昂贵的立体相机装置。
{"title":"Estimating cost of pothole repair from digital images using Stereo Vision and Artificial Neural Network","authors":"Edoghogho Olaye, Eriksson Owraigbo, Nosa Bello","doi":"10.58190/ijamec.2024.77","DOIUrl":"https://doi.org/10.58190/ijamec.2024.77","url":null,"abstract":"A significant amount of road maintenance cost goes into pothole repairs. The primary cost factors related to potholes are their size and depth, as larger and thicker potholes incur higher repair costs. However, existing methods for estimating pothole repair in developing countries rely on manual size measurements, which is time consuming, labor intensive, subjective and can lead to poor estimation of repair cost. This paper presents a system that can automatically determine the size of potholes from digital images and estimate the cost of repair. \u0000In this study, the stereo vision method was used to automatically estimate the depths of potholes from digital camera images. A feed-forward backward propagation Artificial Neural Network (ANN) was trained using pothole images acquired using mobile phones. The predicted depths and sizes of the potholes were then used to estimate the quantity of materials required to fill the potholes and subsequently, the cumulative cost of repair. Marking out and manual size measurements were performed for twenty randomly selected potholes in the Ugbowo Campus of the University of Benin, Nigeria. These measurements were compared against the estimated sizes of potholes predicted by the ANN model. A system was developed to automatically compute these material costs and considering other cost components such as transportation, labor, and equipment.\u0000Results obtained showed that the mean errors for depth, width and height estimates were 3.403%, 3.789% and 5.2617% respectively. Consequently, the developed system correctly estimated the cost of repair of the potholes considered in this study. A significant contribution of the paper is the speed and convenience of acquiring pothole data using a mobile phones without the need for on spot assessment of potholes or use of relatively more expensive stereoscopic camera setup.\u0000","PeriodicalId":496101,"journal":{"name":"International Journal of Applied Methods in Electronics and Computers","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-03-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140358254","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Brushless DC (BLDC) motors, which have small volumes, are widely used in many areas from the aviation industry to industrial applications due to their high efficiency and torque. In parallel with the development of technology, the field of use continues to expand with the development of BLDC engine (BLDCM) control strategies and the decrease in control costs. In this thesis study, it is aimed to minimize the observed changes in rotor speed compared to the reference speed. To achieve this, PID parameters were tried to be changed simultaneously with fuzzy control techniques, taking the error value as a reference. The control system of the BLDC engine was designed in the MATLAB/Simulink environment. In the simulation, the operating stability of classical PID and PID with updated fuzzy-based parameters on two engines with the same features was compared at different speeds. As a result of the research, it was concluded that the correction of the speed observed in the rotor of the PID-controlled motor, whose fuzzy logic-based coefficients were updated, based on the reference speed was more stable and the percentage of exceedance for the reference value was lower, compared to the classical PID controlled motor.
{"title":"BLDC Motor speed control with dynamic adjustment of PID coefficients: Comparison of fuzzy and classic PID","authors":"Selahattin Guntay, Ismail Saritas","doi":"10.58190/ijamec.2024.80","DOIUrl":"https://doi.org/10.58190/ijamec.2024.80","url":null,"abstract":"Brushless DC (BLDC) motors, which have small volumes, are widely used in many areas from the aviation industry to industrial applications due to their high efficiency and torque. In parallel with the development of technology, the field of use continues to expand with the development of BLDC engine (BLDCM) control strategies and the decrease in control costs. In this thesis study, it is aimed to minimize the observed changes in rotor speed compared to the reference speed. To achieve this, PID parameters were tried to be changed simultaneously with fuzzy control techniques, taking the error value as a reference. The control system of the BLDC engine was designed in the MATLAB/Simulink environment. In the simulation, the operating stability of classical PID and PID with updated fuzzy-based parameters on two engines with the same features was compared at different speeds. As a result of the research, it was concluded that the correction of the speed observed in the rotor of the PID-controlled motor, whose fuzzy logic-based coefficients were updated, based on the reference speed was more stable and the percentage of exceedance for the reference value was lower, compared to the classical PID controlled motor.\u0000\u0000","PeriodicalId":496101,"journal":{"name":"International Journal of Applied Methods in Electronics and Computers","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140375932","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Selecting the appropriate honey harvesting method is crucial for sustainable beekeeping and optimal honey production. The use of primitive harvesting methods can lead to the death of bees and a decrease in honey yield. This study aims to address the issue of detecting and classifying young larvae on honeycombs. However, the area where young larvae are found is limited compared to other areas. In this study, the dataset obtained from honeycombs was imbalanced, which has used the Synthetic Minority Oversampling TEchnique (SMOTE) algorithm to balance it. The SMOTE algorithm is a synthetic data generation method. The balanced dataset was then used for classification processes with k-Nearest Neighbors algorithm (k-NN), Decision Trees, and Support Vector Machines. The evaluation of the classification results included the F1-Score, G-Mean, and AUC metrics. The results showed that the classification of the dataset balanced with synthetic data was more successful.
{"title":"Boosting the classification success in imbalanced data of bee larva cells","authors":"Serkan Özgün, M. A. Şahman","doi":"10.58190/ijamec.2024.78","DOIUrl":"https://doi.org/10.58190/ijamec.2024.78","url":null,"abstract":"Selecting the appropriate honey harvesting method is crucial for sustainable beekeeping and optimal honey production. The use of primitive harvesting methods can lead to the death of bees and a decrease in honey yield. This study aims to address the issue of detecting and classifying young larvae on honeycombs. However, the area where young larvae are found is limited compared to other areas. In this study, the dataset obtained from honeycombs was imbalanced, which has used the Synthetic Minority Oversampling TEchnique (SMOTE) algorithm to balance it. The SMOTE algorithm is a synthetic data generation method. The balanced dataset was then used for classification processes with k-Nearest Neighbors algorithm (k-NN), Decision Trees, and Support Vector Machines. The evaluation of the classification results included the F1-Score, G-Mean, and AUC metrics. The results showed that the classification of the dataset balanced with synthetic data was more successful.\u0000\u0000","PeriodicalId":496101,"journal":{"name":"International Journal of Applied Methods in Electronics and Computers","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140374511","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Researches on Unmanned Aerial Vehicles (UAVs) have been recently attracting considerable interest in the field of control theory applications. They are used in a wide range of areas thanks to having the potential of high manoeuvrability, hovering and flying, taking off and landing capabilities. However, to maintain robust control action towards changing conditions of the system is not an easy matter since quadrotor UAVs are highly unstable systems with high precision. Therefore, the main purpose of this study is to control a quadrotor UAV by using a proposed multi-input single-output (MISO) fuzzy-logic controller that ensures robustness if model parameters and trajectory change. For that reason, a 2-dimensional 3 degree-of-freedom quadrotor was used in this study to better evaluate the performance of proposed controller on UAVs. Afterwards, numerical analysis was performed and the findings were analysed. Consequently, the single most striking observation to emerge from the study is that the satisfactory results have been obtained demonstrating that the proposed fuzzy logic controller has remarkable advantage on the robustness of quadrotor UAVs.
{"title":"Robust fuzzy-logic flight control for unmanned aerial vehicles (UAVs)","authors":"Cengiz Özbek","doi":"10.58190/ijamec.2024.79","DOIUrl":"https://doi.org/10.58190/ijamec.2024.79","url":null,"abstract":"Researches on Unmanned Aerial Vehicles (UAVs) have been recently attracting considerable interest in the field of control theory applications. They are used in a wide range of areas thanks to having the potential of high manoeuvrability, hovering and flying, taking off and landing capabilities. However, to maintain robust control action towards changing conditions of the system is not an easy matter since quadrotor UAVs are highly unstable systems with high precision. Therefore, the main purpose of this study is to control a quadrotor UAV by using a proposed multi-input single-output (MISO) fuzzy-logic controller that ensures robustness if model parameters and trajectory change. For that reason, a 2-dimensional 3 degree-of-freedom quadrotor was used in this study to better evaluate the performance of proposed controller on UAVs. Afterwards, numerical analysis was performed and the findings were analysed. Consequently, the single most striking observation to emerge from the study is that the satisfactory results have been obtained demonstrating that the proposed fuzzy logic controller has remarkable advantage on the robustness of quadrotor UAVs.\u0000\u0000","PeriodicalId":496101,"journal":{"name":"International Journal of Applied Methods in Electronics and Computers","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140376239","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Rapid improvements in the area of multi-robot control algorithms pave the way to design and implement robotic swarms to deal with sophisticated tasks including intelligent object transportation systems. It is crucial to manage the structure of the numerous robots to behave like a whole body for task accomplishment. The leader-follower formation control approach offers a simple and reliable way of keeping the swarm formation in appropriate limits to cope with challenging tasks. Autonomous object transportation with multi-robot systems enjoy the benefits of the leader-follower formation control approach. However, most of the developed transportation systems achieve the task by locating the load onto the robots or by pushing the load in the means of a physical contact. These approaches may lead to a hardware or payload damage due to heavy loads or physical contacts respectively. In this study, a novel non-contact object delivery system is introduced for eliminating the drawbacks of physical contact between the robots and the payload. Permanent magnets are used for propulsion of the payload located on a cart with passive casters. The stability of the proposed multi-robot system is satisfied by a formation controller using potential functions method augmented with a cornering action sub-controller. The simulation results verify the effectiveness of the proposed system during a straight motion and cornering with the root mean square values of the distance between the robots as 1.46 × 10-4 [m] and 0.065 [m] respectively.
{"title":"A Non-Contact Object Delivery System Using Leader-Follower Formation Control for Multi-Robots","authors":"Halil İbrahim Dokuyucu, Nurhan Gürsel Özmen","doi":"10.58190/ijamec.2023.40","DOIUrl":"https://doi.org/10.58190/ijamec.2023.40","url":null,"abstract":"Rapid improvements in the area of multi-robot control algorithms pave the way to design and implement robotic swarms to deal with sophisticated tasks including intelligent object transportation systems. It is crucial to manage the structure of the numerous robots to behave like a whole body for task accomplishment. The leader-follower formation control approach offers a simple and reliable way of keeping the swarm formation in appropriate limits to cope with challenging tasks. Autonomous object transportation with multi-robot systems enjoy the benefits of the leader-follower formation control approach. However, most of the developed transportation systems achieve the task by locating the load onto the robots or by pushing the load in the means of a physical contact. These approaches may lead to a hardware or payload damage due to heavy loads or physical contacts respectively. In this study, a novel non-contact object delivery system is introduced for eliminating the drawbacks of physical contact between the robots and the payload. Permanent magnets are used for propulsion of the payload located on a cart with passive casters. The stability of the proposed multi-robot system is satisfied by a formation controller using potential functions method augmented with a cornering action sub-controller. The simulation results verify the effectiveness of the proposed system during a straight motion and cornering with the root mean square values of the distance between the robots as 1.46 × 10-4 [m] and 0.065 [m] respectively.","PeriodicalId":496101,"journal":{"name":"International Journal of Applied Methods in Electronics and Computers","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135081535","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Textile production has an important share in the Turkish economy. One of the common problems in textile factories in Turkey is fabric texture defects that may occur due to textile machinery. The faulty production of the fabric adversely affects the company's economy and prestige. Many methods have been developed to achieve high accuracy in detecting defects in fabric. The aim of this study is to compare the performance of the models using the new dataset and deep learning models. The findings have determined that the Seresnet152d model, which is one of the transfer learning models, can classify with 95.38% accuracy on the generated dataset. Moreover, the majority voting gives 95.58% accuracy rate. In order to achieve high accuracy in the future, it is planned to optimize the parameters of the models used in the study with the help of swarm-oriented optimization algorithms.
{"title":"Ensemble learning application for textile defect detection","authors":"Okan Guder, Sahin Isik, Yildiray Anagun","doi":"10.58190/ijamec.2023.41","DOIUrl":"https://doi.org/10.58190/ijamec.2023.41","url":null,"abstract":"Textile production has an important share in the Turkish economy. One of the common problems in textile factories in Turkey is fabric texture defects that may occur due to textile machinery. The faulty production of the fabric adversely affects the company's economy and prestige. Many methods have been developed to achieve high accuracy in detecting defects in fabric. The aim of this study is to compare the performance of the models using the new dataset and deep learning models. The findings have determined that the Seresnet152d model, which is one of the transfer learning models, can classify with 95.38% accuracy on the generated dataset. Moreover, the majority voting gives 95.58% accuracy rate. In order to achieve high accuracy in the future, it is planned to optimize the parameters of the models used in the study with the help of swarm-oriented optimization algorithms.","PeriodicalId":496101,"journal":{"name":"International Journal of Applied Methods in Electronics and Computers","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135082554","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In the era of the 21st century, Deep CNN has proven its potential in crop and fruit disease classification and detection. Diseases have a ruinous effect on the quality and gross production of yields, which is related to the world economy. Proper identification of diseases at early stages may save yields from damage. CNN-based disease identification can detect the disease at the actual extent at a low cost with minimum expert manpower and labor. Strawberry is considered a functional food, that has a lot of health benefits for the human body. In this study, pre-trained weight ResNet models ResNet50, ResNet101, and ResNet152 architectures are used via the transfer learning features of CNN. Only the classifier of the models is getting updated during training. The Strawberry leaf images are used in this study from the PlantVillage dataset where both classes are balanced in terms of the number of images in each class. Among the three ResNet architectures, ResNet50 outperforms the other ResNet models achieving 88% classification accuracy during the testing period. The ResNet101 and ResNet152 models show 82% and 80% accuracy during the testing period, respectively.
{"title":"ResNet for Leaf-based Disease Classification in Strawberry Plant","authors":"Pranajit Kumar Das, Subarna Sarker Rupa","doi":"10.58190/ijamec.2023.42","DOIUrl":"https://doi.org/10.58190/ijamec.2023.42","url":null,"abstract":"In the era of the 21st century, Deep CNN has proven its potential in crop and fruit disease classification and detection. Diseases have a ruinous effect on the quality and gross production of yields, which is related to the world economy. Proper identification of diseases at early stages may save yields from damage. CNN-based disease identification can detect the disease at the actual extent at a low cost with minimum expert manpower and labor. Strawberry is considered a functional food, that has a lot of health benefits for the human body. In this study, pre-trained weight ResNet models ResNet50, ResNet101, and ResNet152 architectures are used via the transfer learning features of CNN. Only the classifier of the models is getting updated during training. The Strawberry leaf images are used in this study from the PlantVillage dataset where both classes are balanced in terms of the number of images in each class. Among the three ResNet architectures, ResNet50 outperforms the other ResNet models achieving 88% classification accuracy during the testing period. The ResNet101 and ResNet152 models show 82% and 80% accuracy during the testing period, respectively.","PeriodicalId":496101,"journal":{"name":"International Journal of Applied Methods in Electronics and Computers","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135081538","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The acceleration of data production and consumption due to the transition to an information society and industrial revolutions has had a significant impact on the expansion of the global economy. The emergence of Industry 4.0 has led to the adoption of various technologies, including blockchain, which is known for its potential to transform different domains through its solutions. This is particularly relevant in the context of data governance. Thus, blockchain technology has the potential to enhance the sustainability of diverse industries. Sustainability is a crucial concept that refers to the capacity to meet the requirements of the current generation without compromising the ability of future generations to do so. The integration of blockchain technology across diverse industries holds the potential to greatly improve sustainability efforts. The objective of this study is to assess the relationship between blockchain technology and sustainability through a descriptive review of literature utilizing the latent semantic analysis topic modeling and clustering method, which is a social spider optimization technique. This study focuses on analyzing the impact of blockchain technologies on the sustainability sector. A corpus of 1069 papers has been sourced from the Scopus database. The results underscore the significance of cybersecurity, supply chain management, and the circular economy in the extant academic literature. The broad recognition of the supply chain domain's importance is evident in its application of blockchain technology and adherence to the sustainability principle. The present research focuses on the analysis and assessment of topics pertaining to traceability, cyber security, circular economy, energy, and transparency.
{"title":"An analysis of the integration of sustainability concepts into blockchain technology","authors":"Nazmiye Eligüzel","doi":"10.58190/ijamec.2023.43","DOIUrl":"https://doi.org/10.58190/ijamec.2023.43","url":null,"abstract":"The acceleration of data production and consumption due to the transition to an information society and industrial revolutions has had a significant impact on the expansion of the global economy. The emergence of Industry 4.0 has led to the adoption of various technologies, including blockchain, which is known for its potential to transform different domains through its solutions. This is particularly relevant in the context of data governance. Thus, blockchain technology has the potential to enhance the sustainability of diverse industries. Sustainability is a crucial concept that refers to the capacity to meet the requirements of the current generation without compromising the ability of future generations to do so. The integration of blockchain technology across diverse industries holds the potential to greatly improve sustainability efforts. The objective of this study is to assess the relationship between blockchain technology and sustainability through a descriptive review of literature utilizing the latent semantic analysis topic modeling and clustering method, which is a social spider optimization technique. This study focuses on analyzing the impact of blockchain technologies on the sustainability sector. A corpus of 1069 papers has been sourced from the Scopus database. The results underscore the significance of cybersecurity, supply chain management, and the circular economy in the extant academic literature. The broad recognition of the supply chain domain's importance is evident in its application of blockchain technology and adherence to the sustainability principle. The present research focuses on the analysis and assessment of topics pertaining to traceability, cyber security, circular economy, energy, and transparency.","PeriodicalId":496101,"journal":{"name":"International Journal of Applied Methods in Electronics and Computers","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135081542","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Breast cancer is one of the cancer types with a high mortality rate worldwide. Early diagnosis is of great importance to reduce this mortality rate. Computer-aided early diagnosis systems enable doctors to make more precise and faster decisions. The Mammographic Image Analysis Society (MIAS) dataset was used in this study. The breast area was selected by masking in mammography images. The number of images was increased using data augmentation techniques. Mammography images were classified as normal, benign and malignant using four different ResNet models. The highest classification accuracy was achieved by using ResNet18 model with 93.83%. The accuracies obtained with ResNet50, ResNet101 and ResNet152 were 87.24%, 87.44% and 91.25% respectively.
{"title":"Benchmarking of ResNet models for breast cancer diagnosis using mammographic images","authors":"Hasan Serdar Macit, Kadir Sabanci","doi":"10.58190/ijamec.2023.39","DOIUrl":"https://doi.org/10.58190/ijamec.2023.39","url":null,"abstract":"Breast cancer is one of the cancer types with a high mortality rate worldwide. Early diagnosis is of great importance to reduce this mortality rate. Computer-aided early diagnosis systems enable doctors to make more precise and faster decisions. The Mammographic Image Analysis Society (MIAS) dataset was used in this study. The breast area was selected by masking in mammography images. The number of images was increased using data augmentation techniques. Mammography images were classified as normal, benign and malignant using four different ResNet models. The highest classification accuracy was achieved by using ResNet18 model with 93.83%. The accuracies obtained with ResNet50, ResNet101 and ResNet152 were 87.24%, 87.44% and 91.25% respectively.","PeriodicalId":496101,"journal":{"name":"International Journal of Applied Methods in Electronics and Computers","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135081710","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Biomechanical analysis using deep learning has been increasingly used in recent studies to identify human activity. Wearable sensor data from inertial measurement units (IMUs) is widely used for recognizing human activity, but has several drawbacks owing to its high volume and diversity. To overcome these issues, the time-domain and power spectral characteristics of IMU data can be extracted using digital signal processing (DSP) methods. Our research aimed to investigate time-frequency analysis (TFA) methods for classifying the spatio-temporal gait characteristics of physical walking performed by healthy subjects. In this study, open-source biomechanical sensor signal dataset was used. The DSP step was first carried out by segmenting IMU data from the four body segments of 22 healthy subjects, and then by applying Continuous Wavelet Transform (CWT) and Short Time Fourier Transform (STFT) methods. Moreover, the resultants of linear accelerometer signals were applied in a similar manner. The image datasets obtained from this step were applied to a deep convolutional neural network (CNN) model to classify human walking speed (WS) into three classes: fast, normal, and slow. The performance of the 2D-CNN model and the impact of DSP methods using IMU data were evaluated. In conclusion, the highest test classification results demonstrated that STFT-all (85.9%), CWT-all (79.3%), and CWT-resAcc (76.3%) based CNN models present a remarkably precise and easy-to-implement classification problem, with the highest test accuracy, when all IMU channels are subjected to STFT. The classification accuracies of 2D-CNN models were compared to commonly used ML models. The Deep CNN model is a useful gait evaluation tool for healthy subjects. Furthermore, it can enable the diagnosis and phase assessment of gait abnormalities and detect gait biomarkers in rehabilitative wearables.
{"title":"A deep learning approach for human gait recognition from time-frequency analysis images of inertial measurement unit signal","authors":"Hacer Kuduz, Fırat Kaçar","doi":"10.58190/ijamec.2023.44","DOIUrl":"https://doi.org/10.58190/ijamec.2023.44","url":null,"abstract":"Biomechanical analysis using deep learning has been increasingly used in recent studies to identify human activity. Wearable sensor data from inertial measurement units (IMUs) is widely used for recognizing human activity, but has several drawbacks owing to its high volume and diversity. To overcome these issues, the time-domain and power spectral characteristics of IMU data can be extracted using digital signal processing (DSP) methods. Our research aimed to investigate time-frequency analysis (TFA) methods for classifying the spatio-temporal gait characteristics of physical walking performed by healthy subjects. In this study, open-source biomechanical sensor signal dataset was used. The DSP step was first carried out by segmenting IMU data from the four body segments of 22 healthy subjects, and then by applying Continuous Wavelet Transform (CWT) and Short Time Fourier Transform (STFT) methods. Moreover, the resultants of linear accelerometer signals were applied in a similar manner. The image datasets obtained from this step were applied to a deep convolutional neural network (CNN) model to classify human walking speed (WS) into three classes: fast, normal, and slow. The performance of the 2D-CNN model and the impact of DSP methods using IMU data were evaluated. In conclusion, the highest test classification results demonstrated that STFT-all (85.9%), CWT-all (79.3%), and CWT-resAcc (76.3%) based CNN models present a remarkably precise and easy-to-implement classification problem, with the highest test accuracy, when all IMU channels are subjected to STFT. The classification accuracies of 2D-CNN models were compared to commonly used ML models. The Deep CNN model is a useful gait evaluation tool for healthy subjects. Furthermore, it can enable the diagnosis and phase assessment of gait abnormalities and detect gait biomarkers in rehabilitative wearables.","PeriodicalId":496101,"journal":{"name":"International Journal of Applied Methods in Electronics and Computers","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135081533","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：