Browsing by Author "Saminu, Sani"
Now showing 1 - 20 of 53
Results Per Page
Sort Options
Item Analysis And Classification Of Motor Imagery Using Deep Neural Network(institution of Applied Materials and Technology Society with the cooperation of Faculty of Engineering, Universitas Riau, Pekanbaru, Indonesia, 2021-05-25) Ahmad, Isah Salim; Zhang, Shuai; Saminu, Sani; Isselmou; Musa, Jamilu Maaruf; Javaid, Imran; KAMHI, SOUHA; KULSUM, UMMAYMotor imagery based on brain-computer interface (BCI) has aĨracted important research aĨention despite its difficulty. It plays a vital role in human cognition and helps in making the decision. Many researchers use electroencephalogram (EEG) signals to study brain activity with leě and right-hand movement. Deep learning (DL) has been employed for motor imagery (MI). In this article, a deep neural network (DNN) is proposed for classiėcation of leě and right movement of EEG signal using Common Spatial PaĨern (CSP) as feature extraction with standard gradient descent (GD) with momentum and adaptive learning rate LR. (GDMLR), the performance is compared using a confusion matrix, the average classiėcation accuracy is 87%, which is improved as compared with state-of-the-art methods that used different datasets.Item Analysis of Cardiac Beats using Higher Order Spectra(IEEE, 2014-10-29) Karaye, Ibrahim Abdullahi; Saminu, Sani; Özkurt, NalanFor early diagnosis of the heart failures, the electrocardiography (ECG) is the most common method because of its simplicity and cost. Computer based analysis of ECG provides reliable and efficient tools in diagnostics of arrhythmias. With this objective there are lots of studies on automatic and semi-automatic ECG analysis. Like many biosignals, ECG signals are nonlinear in nature, higher order spectral analysis (HOS) is known to be a very good tool for the analysis of nonlinear systems producing good noise immunity. Thus in this study, HOS analysis of ECG signals of normal heart rate, right bundle branch block, paced beat, left bundle block branch and atrial premature beats have been studied in order to reveal the complex dynamics of ECG signals using the tools of nonlinear systems theory. Some of the general characteristics for each of these classes in the bispectrum and bicoherence plot for visual observation have been presented. For the extraction of RR intervals, well known Pan-Tompkins algorithm has been used and three higher order statistical parameters of skewness, kurtosis and variance from these features have been computed. These features with statistical parameters fed into artificial neural network classifier (ANN) and obtained an average accuracy of 94.9%.Item Application of Deep Learning and WT-SST in Localization of Epileptogenic Zone Using Epileptic EEG Signals(MDPI, 2022-05-11) Saminu, Sani; Xu, Guizhi; Zhang, Shuai; Abd El Kader, Isselmou; Jabire, Adamu Halilu; Ahmed, Yusuf Kola; Karaye, Ibrahim Abdullahi; Ahmad, Isah SalimFocal and non-focal Electroencephalogram(EEG) signals have proved to be effective techniques for identifying areas in the brain that are affected by epileptic seizures, known as the epileptogenic zones. The detection of the location of focal EEG signals and the time of seizure occurrence are vital information that help doctors treat focal epileptic seizures using a surgical method. This paper proposed a computer-aided detection (CAD) system for detecting and classifying focal and non-focal EEG signals as the manual process is time-consuming, prone to error, and tedious. The proposed technique employs time-frequency features, statistical, and nonlinear approaches to form a robust features extraction technique. Four detection and classification techniques for focal and non-focal EEG signals were proposed. (1). Combined hybrid features with Support Vector Machine (Hybrid-SVM) (2). Discrete Wavelet Transform with Deep Learning Network (DWT-DNN) (3). Combined hybrid features with DNN (Hybrid-DNN) as an optimized DNN model. Lastly, (4). A newly proposed technique using Wavelet Synchrosqueezing Transform-Deep Convolutional Neural Network (WTSST-DCNN). Prior to feeding the features to classifiers, statistical analyses, including t-tests, were deployed to obtain relevant and significant features at each approach. The proposed feature extraction technique and classification proved effective and suitable for smart Internet of Medical Things (IoMT) devices as performance parameters of accuracy, sensitivity, and specificity are higher than recently related works with a value of 99.7%, 99.5%, and 99.7% respectively.Item APPLICATION OF SUPPORT VECTOR REGRESSION MODELLING FOR THE PREDICTION OF IMPACT ATTENUATION OF 3D PRINTED HIP PROTECTORS(Faculty of Engineering and Technology, University of Ilorin, 2023-06) Suleiman Abimbola, Yahaya; Muniru, I.O.; Saminu, Sani; Ibitoye, M.O.; Ajibola, T.M.; Jilantikiri, L. J.; Ripin, Z.M.; Ridzwan, M.I.Z.3D printed thermoplastic polyurethanes of different shore hardness were used to make hip protectors for the prevention of osteoporotic hip fracture, which was then tested. The result was used to develop a support vector regression model to estimate the effect of the protector shore hardness, shell thickness, and infill density on the impact attenuation capacity at different energy levels. The results from the model show that the impact attenuation ability of a hip protector is significantly dependent on the infill density of the hip protector and its shore hardness. Excellent agreement was found between the model results and test results.Item Applications of Artificial Intelligence in Automatic Detection of Epileptic Seizures Using EEG Signals: A Review(BON VIEW PUBLISHING PTE. LTD, 2022-07) Saminu, Sani; Xu, Guizhi; Zhang, Shuai; Abd El Kader, Isselmou; Aliyu, Hajara Abdulkarim; Jabire, Adamu Halilu; Ahmed, Yusuf Kola; Adamu, Mohammed JajereCorrectly interpreting an electroencephalogram signal with high accuracy is a tedious and time-consuming task that may take several years of manual training due to its complexity, noisy, non-stationarity, and nonlinear nature. To deal with the vast amount of data and recent challenges of meeting the requirements to develop low cost, high speed, low complexity smart internet of medical things computer-aided devices (CAD), artificial intelligence (AI) techniques which consist of machine learning and deep learning (DL) play a vital role in achieving the stated goals. Over the years, machine learning techniques have been developed to detect and classify epileptic seizures. But until recently, DL techniques have been applied in various applications such as image processing and computer visions. However, several research studies have turned their attention to exploring the efficacy of DL to overcome some challenges associated with conventional automatic seizure detection techniques. This article endeavors to review and investigate the fundamentals, applications, and progress of AI-based techniques applied in CAD system for epileptic seizure detection and characterization. It would help in actualizing and realizing smart wireless wearable medical devices so that patients can monitor seizures before their occurrence and help doctors diagnose and treat them. The work reveals that the recent application of DL algorithms improves the realization and implementation of mobile health in a clinical environment.Item Brain Tumor Detection and Classification by Hybrid CNN-DWA Model Using MR Images(Bentham Science, 2021-02-23) Isselmou, Abd El Kader; Guizhi, Xu; Zhang, Shuai; Saminu, SaniObjective: Medical image processing is an exciting research area. In this paper, we proposed new brain tumor detection and classification model using MR brain images to help the doctors in early detection and classification of the brain tumor with high performance and best accuracy. Materials: The model was trained and validated using five databases, including BRATS2012, BRATS2013, BRATS2014, BRATS2015, and ISLES-SISS 2015. Methods: The advantage of the hybrid model proposed is its novelty that is used for the first time; our new method is based on a hybrid deep convolution neural network and deep watershed auto-encoder (CNN-DWA) model. The method consists of six phases, first phase is input MR images, second phase is preprocessing using filter and morphology operation, third phase is matrix that represents MR brain images, fourth is applying the hybrid CNN-DWA, fifth is brain tumor classification, and detection, while sixth phase is the performance of the model using five values. Results: The novelty of our hybrid CNN-DWA model showed the best results and high performance with accuracy around 98% and loss validation 0, 1. Hybrid model can classify and detect the tumor clearly using MR images; comparing with other models like CNN, DNN, and DWA, we discover that the proposed model performs better than the above-mentioned models. Conclusion: Depending on the better performance of the proposed hybrid model, this helps in developing computer-aided system for early detection of brain tumors and helps the doctors to diagnose the patients better.Item Brain Tumor Detection and Classification on MR Images by a Deep Wavelet Auto‐Encoder Model(MDPI, 2021-08-31) Isselmou, Abd El Kader; Guizhi, Xu; Zhang, Shuai; Saminu, Sani; Javaid, Imran; Ahmad, Isah Salim; KAMHI, SOUHAThe process of diagnosing brain tumors is very complicated for many reasons, including the brain’s synaptic structure, size, and shape. Machine learning techniques are employed to help doctors to detect brain tumor and support their decisions. In recent years, deep learning techniques have made a great achievement in medical image analysis. This paper proposed a deep wavelet autoencoder model named “DWAE model”, employed to divide input data slice as a tumor (abnor‐ mal) or no tumor (normal). This article used a high pass filter to show the heterogeneity of the MRI images and their integration with the input images. A high median filter was utilized to merge slices. We improved the output slices’ quality through highlight edges and smoothened input MR brain images. Then, we applied the seed growing method based on 4‐connected since the thresholding cluster equal pixels with input MR data. The segmented MR image slices provide two two‐layer using the proposed deep wavelet auto‐encoder model. We then used 200 hidden units in the first layer and 400 hidden units in the second layer. The softmax layer testing and training are performed for the identification of the MR image normal and abnormal. The contribution of the deep wavelet auto‐encoder model is in the analysis of pixel pattern of MR brain image and the ability to detect and classify the tumor with high accuracy, short time, and low loss validation. To train and test the overall performance of the proposed model, we utilized 2500 MR brain images from BRATS2012, BRATS2013, BRATS2014, BRATS2015, 2015 challenge, and ISLES, which consists of normal and ab‐ normal images. The experiments results show that the proposed model achieved an accuracy of 99.3%, loss validation of 0.1, low FPR and FNR values. This result demonstrates that the proposed DWAE model can facilitate the automatic detection of brain tumors.Item Brain Tumor identification by Convolution Neural Network with Fuzzy C-mean Model Using MR Brain Images(North Atlantic University Union, 2020-12-29) Isselmou, Abd El Kader; Guizhi, Xu; Zhang, Shuai; Saminu, Sani; Javaid, Imran; Ahmad, Isah SalimMedical image computing techniques are essential in helping the doctors to support their decision in the diagnosis of the patients. Due to the complexity of the brain structure, we choose to use MR brain images because of their quality and the highest resolution. The objective of this article is to detect brain tumor using convolution neural network with fuzzy c-means model, the advantage of the proposed model is the ability to achieve excellent performance using accuracy, sensitivity, specificity, overall dice and recall values better than the previous models that are already published. In addition, the novel model can identify the brain tumor, using different types of MR images. The proposed model obtained accuracy with 98%.Item Brain Tumor Identification by Hybrid CNN-SWT Model(Bentham Science Publishers Ltd, 2022-05-24) Abd El Kader, Isselmou; Xu, Guizhi; Zhang, Shuai; Saminu, Sani; Javaid, Imran; Ahmad, Isah Salim; Kamhi, SouhaObjective: Detecting brain tumor using the segmentationtechnique is a big challenge for researchers and takes a long time inmedical image processing. Magnetic resonance image analysis techniquesfacilitate the accurate detection of tissues and abnormal tumors in thebrain. The size of a brain tumor can vary with the individual and thespecifics of the tumor. Radiologists face great difficulty in diagnosing andclassifying brain tumors. Method: This paper proposed a hybrid model-based convolutional neuralnetwork with a stationary wavelet trans-form named “CNN-SWT” tosegment brain tumors using MR brain big data. We utilized 7 layers forclassification in the proposed model that include 3 convolutional and 3ReLU. Firstly, the input MR image is divided into multiple patches, and thenthe central pixel value of each patch is provided to the CNN-SWT. Secondly,the pre-processing stage is per-formed using the mean filter to remove thenoise. Then the convolution neural network-layer approach is utilized tosegment brain tumors. After segmentation, robust feature extraction suchas information-extraction methods is used for the feature extractionprocess. Finally, a CNN-based hybrid scheme based on the stationarywavelet transform technique is used to detect tumors using MR brainimages. Materials: These experiments were obtained using 11500 MR brain imagesdata from the hospital national of oncology. Results: It was proved that the proposed hybrid achieved a highclassification accuracy of (98.7 %) as compared with existing methods. Conclusion: The advantage of the hybrid novelty of the model and theability to detect the tumor area achieved excellent overall performanceusing different values.Item CHARACTERISTIC MODE ANALYSIS OF A STEPPED GRADIENT PLANAR ANTENNA FOR UWB APPLICATION(Department of Electrical Engineering, Ahmadu Bello University, Zaria, 2020-03) Jabire, Adamu Halilu; Saminu, Sani; Abdu, Anas; abel, Noku Amos; Sadiq, Abubakar MuhammadCharacteristic mode technique is employed to gain a physical insight and also to find out the dominant mode of stepped gradient planar antenna without considering the feeding port. The rational of the model further implies that we can consider antenna's shape and feed design as independent steps. The stepped gradient planar antenna is constructed and measured, in which both the measured and simulated agreed on each other in terms of reflection coefficient and voltage standing wave ratio. The miniature stepped gradient planar antenna is appropriate for numerous applications in ultra-wideband (UWB) communication systems from 2.7 to 12GHz and stable radiation pattern at both E and Hfields were attained over the operating frequency band which is suitable for use in UWB systems.Item Circuit Modeling of Dual Band MIMO Diversity Antenna for LTE and X-Band Applications(Universitas Ahmad Dahlan, 2023-09) Abdullahi, Aminu Gambo; Kolawale, S. F.; Saminu, Sani; Danladi, Ali; Jabire, Adamu HaliluThis paper presents a study on developing a dual-band antenna equivalent circuit model for X-Band and LTE applications. MIMO antennas play a crucial role in modern wireless communication systems, and understanding their impedance behavior is essential. This work proposes a dual-band lumped equivalent circuit model, utilizing gradient optimization based on antenna-simulated S-parameters in Advanced Design System (ADS). The four radiating elements of the MIMO antenna are accurately modeled, considering their geometry and the defected ground structure (DGS) effect, which enhances the antenna's isolation and low correlation coefficient (ECC). The calculated lumped equivalent circuit model is validated through rigorous simulation and measurement data, demonstrating consistency with the expected results. The experimental measurements show measured isolation exceeding 20 dB while achieving a maximum realized gain of 5.9 dBi and an efficiency of 87%. The developed model holds promise for improving the design and performance of MIMO antennas for various applications.Item Classification of Cardiac Beats Using Discrete Wavelet Features(Covenant University, Otta Nigeria, 2015-06) Saminu, Sani; Özkurt, NalanWith the growing technology, the tools which continuously monitor the health status of the people are becoming the integral part of our lives. The detection of a cardiac disease or tracking the heart activities for ongoing cardiac conditions is now possible with portable electrocardiography (ECG) monitors. For detection and classification of ECG signals in portable devices, the robust features and efficient classification algorithms are very important. Thus, in this study, a robust feature set based on discrete wavelet transform (DWT) is proposed, and the performance of the classification tools such as artificial neural networks, support vector machines and probabilistic neural networks are compared. After preprocessing, the R peaks are located by the well-known Pan Tompkins algorithm and 200 samples are taken as equivalent R-T interval in the proposed technique. The statistical parameters such as mean, median, standard deviation, maximum, minimum, energy and entropy of DWT coefficients are used as the feature set. The proposed hybrid technique has been tested by classifying three ECG beats as normal, right bundle branch block (Rbbb) and paced beat using the signals from Massachusetts Institute of Technology Beth Israel Hospital (MIT-BIH) arrhythmia database and processed using Matlab 2013 environment. The best accuracy of 99.84% has been obtained by Db4 mother wavelet with artificial neural network as classifier.Item COMPARISON OF DEEP LEARNING ALEXNET AND SUPPORT VECTOR MACHINE TO CLASSIFY SEVERITY OF SICKLE CELL ANEMIA(Gombe State University, 2022-08) Aliyu, Hajara Abdulkarim; IBRAHIM, MUHAMMAD JAMILU; Saminu, Sani; MUHAMMAD, FATIMA ABDULLAHISickle cell anemia (SCA) is a serious hematological blood disorder, where affected patients are frequently hospitalized throughout a lifetime. Most of the patient's life span reduced, and some become addict based on the nature of strong analgesic that is taken by the concern patients, which they all have strong side effects. The existing method of severity classification for SCA patient is done manually through a microscope which is time-consuming, tedious, prone to error, and require a trained hematologist. The affected patient has many cell shapes that show important biomechanical characteristics of patient severity level. The main purpose of the study is to develop an automated severity level classification method of SCA patients by comparing deep learning AlexNet and Support Vector Machine (SVM) to enable present the percentage of each cell present in blood smear image. Hence, having an effective way of classifying the abnormalities present in the SCA disease based on the level of patient severity to give a better insight into managing the concerned patient's life. The study was performed with 182 SCA patients (over 11,000 single RBC images) with 14 classes of abnormalities and a class of normal cells to develop a shape factor quantification and general multiscale shape analysis to classify the patient based on severity level. As a result, it was found that the proposed framework can detect 85.4% abnormalities in SCA patient blood smear in automated manner when compared with Support Vector Machine (SVM) method with 71.9%. Hence, the system classifies the severity of SCA patient automatically and reduce the time and eye stress with performance AlexNet model performance of 95.1% accuracy, 99.1% specificity, and 98.5% precision value.Item Comparison of Wavelet and Filtering Techniques for Denoising ECG Signal(Faculty of Engineering, Bayero University Kano, Nigeria, 2014) Saminu, SaniA proper processing of biomedical signals enhances their physiological and clinical information because they carry vital information about the behaviour of the living systems under study. With the analysis of the Electrocardiogram (ECG) signal it may be possible to predict heart problems and play an important role in diagnosis process or monitor patient recovery after a heart intervention. The quality of this signal is degraded mainly by many sources of noise such as power line interference (PLI), baseline drift, muscle contraction noise etc. Present work deals with the design of filter banks based on the discrete wavelet transform (DWT) as well as design of low pass Chebyshev Type I and Butterworth filters using FDA tool in MATLAB environment. De-noised ECG signal is compared with original signal using Mean Square Error (MSE). Results show that denoising schemes involving wavelet domains are able to reduce noise from ECG signals more accurately and consistently with mse of 0.0012 in comparison to noise reduction algorithms in filtering technique which has mse of 0.0799 and 0.1814 for Chebyshev and Butterworth filters respectively.Item Convolutional Neural Networks Model for Emotion Recognition Using EEG Signal(North Atlantic University Union, 2021-04-29) Ahmad, Isah Salim; Zhang, Shuai; WANG, LINGYUE; Saminu, Sani; Isselmou, Abd El Kader; CAI, ZILIANG; Javaid, Imran; KAMHI, SOUHA; KULSUM, UMMAYA Brain-computer interface (BCI) using an electroencephalogram (EEG) signal has a great attraction in emotion recognition studies due to its resistance to humans’ deceptive actions. This is the most significant advantage of brain signals over speech or visual signals in the emotion recognition context. A major challenge in EEG-based emotion recognition is that a lot of effort is required for manually feature extractor, EEG recordings show varying distributions for different people and the same person at different time instances. The Poor generalization ability of the network model as well as low robustness of the recognition system. Improving algorithms and machine learning technology helps researchers to recognize emotion easily. In recent years, deep learning (DL) techniques, specifically convolutional neural networks (CNNs) have made excellent progress in many applications. This study aims to reduce the manual effort on features extraction and improve the EEG signal single model’s emotion recognition using convolutional neural network (CNN) architecture with residue block. The dataset is shuffle, divided into training and testing, and then fed to the model. DEAP dataset has class 1, class 2, class 3, and class 4 for both valence and arousal with an accuracy of 90.69%, 91.21%, 89.66%, 93.64% respectively, with a mean accuracy of 91.3%. The negative emotion has the highest accuracy of 94.86% fellow by neutral emotion with 94.29% and positive emotion with 93.25% respectively, with a mean accuracy of 94.13% on the SEED dataset. The experimental results indicated that CNN Based on residual networks can achieve an excellent result with high recognition accuracy, which is superior to most recent approaches.Item A crossed-polarized four port MIMO antenna for UWB communication(Elsevier, 2022-12-29) Jabire, Adamu Halilu; Salisu, Sani; Saminu, Sani; Adamu, Mohammed Jajere; Hussein, Mousa I.This paper presents a compact, crossed-polarized, ultra-wideband (UWB) four-ports multiple- input-multiple-output (MIMO) printed antenna. The proposed antenna is constructed from four microstrip circular patch elements fed by a 50-Ω microstrip line. Two metamaterial cell elements, in the form of a rectangular concentric double split ring resonator (SRR), are placed at the upper plane of the substrates for bandwidth improvement and isolation enhancement. The ultra- wideband frequency response is achieved using a defective ground plane. Surface current flow between the antenna’s four elements is limited to ensure maximum isolation. The four-port MIMO system is designed with orthogonal antenna elements orientation on an FR4 substrate with a loss tangent of 0.02 and an overall size of 30 mm ×30 mm ×1.6 mm. Such orientation resulted in less than 17dB port-to-port isolation and an impedance bandwidth of 148% (3.1–12 GHz). The proposed UWB-MIMO antenna achieved a maximum realized gain of 6.2dBi with an efficiency of 87%. The measured and simulated results are in good agreement over the operating frequency band (3.1–12 GHz). The results also provide overall good diversity performance with the TARC <10 dB, ECC <0.001, DG >9.9, MEG <3 dB and CCL <0.1. The proposed antenna is well- suited for applications in WLAN, WIMAX and GPRs.Item Deep learning algorithm for brain tumor detection and analysis using MR brain images(ACM Digital Library, 2019-07-01) Isselmou, Abd El Kader; Guizhi, Xu; Zhang, Shuai; Saminu, Sani; Javaid, ImranMedical image processing paly a good role in helping the radiologists and facility patients diagnosis, the aims of this paper is created deep learning algorithm to detect brain tumor using magnetic resonance brain images and analysis the performance of algorithm based on different values, accuracy, sensitivity, specificity, ndice, nJaccard coeff and recall values. The significance of convolution neural network (CNN) it’s the ability to detect brain clearly with high performance. We propose framework is successfully tested on data source on magnetic resonance brain images of the patients suffering from different brain tumor types reaching a Dice similarity 86,785% and high accuracy 98, 33%.Item Deep Learning Based on CNN for Emotion Recognition Using EEG Signal(WSEAS, 2021-04-14) Ahmad, Isah Salim; Zhang, Shuai; Saminu, Sani; WANG, LINGYUE; Isselmou, Abd El Kader; CAI, ZILIANG; Javaid, Imran; KAMHI, SOUHA; KULSUM, UMMAYEmotion recognition based on brain-computer interface (BCI) has attracted important research attention despite its difficulty. It plays a vital role in human cognition and helps in making the decision. Many researchers use electroencephalograms (EEG) signals to study emotion because of its easy and convenient. Deep learning has been employed for the emotion recognition system. It recognizes emotion into single or multi-models, with visual or music stimuli shown on a screen. In this article, the convolutional neural network (CNN) model is introduced to simultaneously learn the feature and recognize the emotion of positive, neutral, and negative states of pure EEG signals single model based on the SJTU emotion EEG dataset (SEED) with ResNet50 and Adam optimizer. The dataset is shuffle, divided into training and testing, and then fed to the CNN model. The negative emotion has the highest accuracy of 94.86% fellow by neutral emotion with 94.29% and positive emotion with 93.25% respectively. With average accuracy of 94.13%. The results showed excellent classification ability of the model and can improve emotion recognition.Item Design and Construction of a Portable Electronic Sleep Inducer for Low Resource Settings(Faculty of Engineering, FUOYE, Nigeria, 2020-09) Ahmed, Yusuf Kola; Zubair, Abdul R.; Saminu, Sani; Akande, Kareem A.; Afolayan, Mubarak A.; Afonja, Awawu A.Good quality restful sleep is indispensable to mental and physical health. However, pressure due to busy life style, work and sometimes physiological factors have placed constraints on adequate and healthy sleep pattern leading to several sleep disorders such as insomnia, sleep apnea and restless leg syndrome. Sleep disorder affects the quality of life of such patients as it grossly reduces efficiency at work and leads to poor mental and physical health. Available drugs to treat this disorder are addictive with strong adverse effects, while existing devices to provide intervention are very expensive. Hence, the development of an affordable, portable electronic sleep inducer with display unit is presented. It uses geomagnetic property of the earth coupled with electromagnetic wave induction to stimulate sleep. The signal frequency was generated by IC4047 coupled with Arduino Uno and ATmega 328p for device control. The output of this electronic sleep inducer is found to satisfactorily produce 5.89 Hz and 3.58 Hz for theta and delta waves respectively, needed to induce sleep. It consumes less power and it is rechargeable.Item Design and Development of a Hybrid Eye and Mobile Controlled Wheelchair Prototype using Haar cascade Classifier: A Proof of Concept(Springer Nature Switzerland, 2023-08) Ahmed, Yusuf Kola; Suleiman, Taofik Ahmed; Saminu, Sani; Danmusa1, Nasir Ayuba; Salahudeen, Kafilat Atinuke; Zubair, Abdul Rasak; Adelodun, Abdulwasiu BolakaleAccording to the wheelchair foundation, about 1.86% of the world’s population requires a functional wheelchair. Most of these wheelchairs have manual control systems which puts millions of people with total paralyzes (total loss of muscle control including the head) at a disadvantage. However, the majority of those who suffer from muscular and neurological disorders still retain the ability to move their eyes. Hence the concept of eye-controlled wheelchair. This paper focused on the design and development of a hybrid control system (eye and mobile interface) for a wheelchair prototype as a proof of concept. The systemwas implemented using the pre-trained Haar cascade ML classifier in open CV. Focus was shifted from high accuracy common to lab-based studies to deployment and power consumption which are critical to usability. The system consists of a motor chassis that takes the place of a wheelchair, a raspberry pi4 module which acts as a mini-computer for image and information processing, and a laser sensor to achieve obstacle avoidance. The Bluetooth module enables serial communication between the motor chassis and the raspberry pi, while the power supply feeds the raspberry pi and the camera. The system performance evaluation was carried out using obstacle avoidance and navigation tests. An accuracy of 100% and 89% were achieved for obstacle avoidance and navigation, respectively, which shows that the system would be helpful for wheelchair users facing autonomous mobility issues.
- «
- 1 (current)
- 2
- 3
- »