Taiwan Association of Engineering and Technology Innovation: E-Journals
Not a member yet
887 research outputs found
Sort by
A Convolutional Neural Network for Automatic Brain Tumor Detection
Get PDFMagnetic resonance imaging (MRI) combined with artificial intelligence (AI) algorithms to detect brain tumors is one of the important medical applications. In this study, a Convolutional neural network (CNN) model is proposed to detect meningioma and pituitary, which was tested with a dataset consisting of two categories of tumors with 1,800 MRI images from several persons. The CNN model is trained via a Python library, namely TensorFlow, with an automatic tuning approach to obtain the highest testing accuracy of tumor detection. The CNN model used Python programming language in Google Colab to detect sensitivity, precision, the area under the PR and receiver operating characteristic (ROC), error matrix, and accuracy. The results show that the proposed CNN model has a high performance in the detection of brain tumors. It achieves an accuracy of 95.78% and a weighted average precision of 95.82%
Optimization of Weld Parameters in Wire and Arc-Based Directed Energy Deposition of High Strength Low Alloy Steels
Get PDFThis paper aims to investigate the fabrication of high strength low alloy (HSLA) steels by wire and arc-based directed energy deposition (WADED). Firstly, the relationship between the process variables (including the travel speed-V, the current-C, and the voltage-U) and the geometrical characteristics of weld beads (including the bead height (BH), bead width (BW), and melting pool length (MPL)) was investigated. Secondly, the optimal process variables were identified using the desirability approach. The results indicate that voltage-U has the highest impact on BW and MPL, meanwhile the travel speed-V is the most impacting factor on BH. The optimal variables for the WADED process of HSAL steels are V = 0.3 m/min, C = 160 A, and U = 19 V. The component fabricated with the optimal variables is fully dense without spatters and defects, confirming the efficiency of the WADED process for HSLA steels
Improved Preprocessing Strategy under Different Obscure Weather Conditions for Augmenting Automatic License Plate Recognition
Get PDFAutomatic license plate recognition (ALPR) systems are widely used for various applications, including traffic control, law enforcement, and toll collection. However, the performance of ALPR systems is often compromised in challenging weather and lighting conditions. This research aims to improve the effectiveness of ALPR systems in foggy, low-light, and rainy weather conditions using a hybrid preprocessing methodology. The research proposes the combination of dark channel prior (DCP), non-local means denoising (NMD) technique, and adaptive histogram equalization (AHE) algorithms in CIELAB color space. And used the Python programming language comparisons for SSIM and PSNR performance. The results showed that this hybrid approach is not merely robust to a variety of challenging conditions, including challenging weather and lighting conditions but significantly more accurate for existing ALPR systems
Design Optimization of a Capacitive Sensor for Mass Measurement of Nanometer-Sized Exhaust Carbon Particles
Get PDFNanometer-sized carbon particulates generated by incomplete combustion in heavy-duty vehicles are harmful to human health. A high-resolution technique is needed to detect and measure these pollutants. This study aims to optimize a capacitive sensor design for detecting and measuring particulates. Firstly, the effect of design parameters on particulate detection and sensor compliance sensitivity is investigated by using the finite element method. By comparing the simulation results with literature findings for performance validation, the sensor structure is optimized to detect lower particulate concentrations. The simulation result shows that particulate detection sensitivity has linear variations with changes in particulate mass. With optimum electrode spacing and top insulation layer thickness of 5 µm, the sensor can detect a particulate deposition of 0.033 mg/min and generate a maximum capacitance of 581 pF. Since the optimized design can measure particulate deposition at a lower range and with higher sensitivity, it is suitable to be applied to detect nanometer-sized carbon particulates
Development of Mixing and Pressing Processes of Split-Gill Mushroom Spawn Blocks
Get PDFThis study aims to develop the mixing and pressing processes of split-gill mushroom spawn blocks through the development and construction of a semi-automatic mushroom spawn mixing. The developed machine uses a 0.5 hp motor to drive the mixing tank and the press cylinder, which are connected to a 1:60 reduction gear. The results show that the semi-automatic mushroom spawns mixing and pressing machine developed in this study are within the standard ranges, that the split-gill mushroom spawn blocks with an average weight of 598 g, an average height of 10.2 cm, and an average density of 0.33 g/cm3. As for production capacity, manual pressing produced 40 mushroom spawn blocks per hour while the developed machine produced 112 mushroom spawn blocks per hour, which is 2.8 times faster
Fade Lighting Control Method for Visual Comfort and Energy Saving
Get PDFThis study proposes a fade lighting control method to ensure the visual comfort of indoor occupants through gradual illuminance control while saving energy. The illuminance sensor measures the indoor illuminance and calculates the required illuminance for achieving a reference illuminance of 500 Lux. The control illuminance for each lighting is derived based on the required illuminance, and it is confirmed to fall within the threshold range of 20%. The illuminance values and time intervals for fade lighting control are calculated, ensuring that the amount of illuminance adjustment is divided by the size of the threshold range or less. In the performance evaluation, the proposed method (experimental group) was compared with the influence-based control method (control group). The result shows that this fade lighting control method minimizes the visual discomfort of occupants caused by sudden changes in lighting, and the same energy-saving of 11-42% is achieved as the control group
The Control Method for Wavelength-Based CCT of Natural Light Using Warm/Cool White LED
Get PDFReproducing circadian patterns of natural light through lighting requires technology that can control correlated color temperature (CCT) and short wavelength ratio (SWR) simultaneously. This study proposes a method for controlling wavelength-based CCT of natural light using LED light sources. First, the spectral power distribution (SPD) of each channel of the test lighting (two-channel LED lighting with warm white and cool white) is identified through actual measurement. Next, CCT and SWR are calculated based on the additive mixing of SPD using the mixing ratio from the measured SPD. Finally, the regression equations for mixing ratio-CCT and mixing ratio-SWR are derived through regression analysis. These equations are then utilized to implement a wavelength-based CCT control algorithm. For performance and evaluation purposes, natural light reproduction experiments were conducted, achieving a mean error of 94.5K for CCT and 1.5% for SWR
Development of a New Ground Motion Model for a Peninsular Indian Rock Site
Get PDFThe ground motion model (GMM) plays a vital role in the generation of seismic design basis ground motion parameters. Even though many intra-plate GMMs are available, very few of them are based on Peninsular India (PI) region-specific seismological parameters. Hence, it is imperative to develop a GMM using seismological parameters derived from earthquakes in the Peninsular Indian region. In this study, a new GMM is developed for a PI rock site. Due to the scarcity of real earthquakes, artificial earthquake records are simulated to generate a new GMM for PI. The accelerograms of these artificial earthquakes are obtained from the stochastic finite fault simulation technique. Region-specific seismological parameters are obtained from the available PI earthquakes. The generated GMM is compared with other intra-plate GMMs for different earthquake magnitudes. Also, the generated GMM is validated with the Koyna earthquake record and it is observed that the GMM’s predictions are closer to the record
Deep Learning-Based Iris Segmentation Algorithm for Effective Iris Recognition System
Get PDFIn this study, a 19-layer convolutional neural network model is developed for accurate iris segmentation and is trained and validated using five publicly available iris image datasets. An integrodifferential operator is used to create labeled images for CASIA v1.0, CASIA v2.0, and PolyU Iris image datasets. The performance of the proposed model is evaluated based on accuracy, sensitivity, selectivity, precision, and F-score. The accuracy obtained for CASIA v1.0, CASIA v2.0, CASIA Iris Interval, IITD, and PolyU Iris are 0.82, 0.97, 0.9923, 0.9942, and 0.98, respectively. The result shows that the proposed model can accurately predict iris and non-iris regions and thus can be an effective tool for iris segmentation
Short-Term Rainfall Prediction Using Supervised Machine Learning
Get PDFFloods and rain significantly impact the economy of many agricultural countries in the world. Early prediction of rain and floods can dramatically help prevent natural disaster damage. This paper presents a machine learning and data-driven method that can accurately predict short-term rainfall. Various machine learning classification algorithms have been implemented on an Australian weather dataset to train and develop an accurate and reliable model. To choose the best suitable prediction model, diverse machine learning algorithms have been applied for classification as well. Eventually, the performance of the models has been compared based on standard performance measurement metrics. The finding shows that the hist gradient boosting classifier has given the highest accuracy of 91%, with a good F1 value and receiver operating characteristic, the area under the curve score