• Journal of Biomedical Engineering Research
• /
• v.8 no.2
• /
• pp.215-222
• /
• 1987

This paper describes the design of a real time arrhythmia detection system using five variables : heart rate, QS width, morphology, alls deviation, and ST segment. To detect individual variables for four patients, we designed multi-processor system. The results of the analysis derived from simulators and 50 patients are compared with the medical diagnoses. The results show that the analysis was able to detect irregularly occuring arrhythmia which does not show up in routine medical examination.

• Proceedings of the IEEK Conference
• /
• 2002.07b
• /
• pp.1256-1259
• /
• 2002

In this paper, we have proposed license plate recognition system for multi-national vehicle license plate using geometric features along with hybrid and seven segment pattern vectors. In the proposed system, we suggested to find horizontal and vertical relation after going through preparation process with inputted real-time license plate image of Korea and Japan, and then to classify license plate with using characteristic and geometric information of license plates. It classifies the extracted license plate images into letters and numbers, such as local name, local number, classification character and license consecutive numbers, and recognize license plate of Korea and Japan by applying hybrid and seven segments pattern vectors to classified letter and number region. License plate extraction step of the proposed system uses width and length information along with relative rate of Korean and Japanese license plate. Moreover, it exactly segmentation by letters with using each letter and number position information within license plate region, and recognizes Korean and Japanese license plates by applying hybrid and seven segment pattern vectors, containing characteristics related to letter size and movement within segmented letter area. As the result of testing the proposed system in real experiment, it recognized regardless of external lighting conditions as well as classifying license plates by nations, Korea and Japan. We have developed a system, recognizing regardless of inputted structural character of vehicle licenses and external environment.

• Korean Management Science Review
• /
• v.33 no.2
• /
• pp.75-87
• /
• 2016

Quality function deployment (QFD) is a widely adopted customer-oriented product development methodology by analyzing customer requirements. It is a main activity in QFD planning process to determine the optimal values of the technical attributes (TAs) so as to achieve the customer requirements (CRs) from the House of Quality (HoQ). In most of the previous research, all the TAs in QFD are assumed to have either continuous or discrete values. In the real world applications, the continuous TAs and the discrete TAs are often mixed in QFD. In this paper, a mixed integer linear programming model is formulated to obtain the optimal values for the continuous TAs and the discrete TAs in QFD planning as well as Branch and Bound (B and B) algorithm is proposed as the solution approach. Finally, the proposed model and solution approach are illustrated with an office chair under multi-segment market, and the sensitivity analysis is performed to study how the proposed model and its solutions respond to the variation for the two elements which are budget and CRs' weights.

• Smart Structures and Systems
• /
• v.24 no.4
• /
• pp.475-490
• /
• 2019

The detection of structural damage without a priori information on the healthy state is challenging. In order to address the issue, the study presents a baseline-free approach to detect damage in beam structures based on an actual influence line. In particular, a multi-segment function-fitting calculation is developed to extract the actual deflection influence line (DIL) of a damaged beam from bridge responses due to a passing vehicle. An intact basis function based on the measurement position is introduced. The damage index is defined as the difference between the actual DIL and a constructed function related to the intact basis, and the damage location is indicated based on the local peak value of the damage index curve. The damage basis function is formulated by using the detected damage location. Based on the intact and damage basis functions, damage severity is quantified by fitting the actual DIL using the least-square calculation. Both numerical and experimental examples are provided to investigate the feasibility of the proposed method. The results indicate that the present baseline-free approach is effective in detecting the damage of beam structures.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.10
• /
• pp.3668-3684
• /
• 2021

Video action recognition is widely used in video surveillance, behavior detection, human-computer interaction, medically assisted diagnosis and motion analysis. However, video action recognition can be disturbed by many factors, such as background, illumination and so on. Two-stream convolutional neural network uses the video spatial and temporal models to train separately, and performs fusion at the output end. The multi segment Two-Stream convolutional neural network model trains temporal and spatial information from the video to extract their feature and fuse them, then determine the category of video action. Google Xception model and the transfer learning is adopted in this paper, and the Xception model which trained on ImageNet is used as the initial weight. It greatly overcomes the problem of model underfitting caused by insufficient video behavior dataset, and it can effectively reduce the influence of various factors in the video. This way also greatly improves the accuracy and reduces the training time. What's more, to make up for the shortage of dataset, the kinetics400 dataset was used for pre-training, which greatly improved the accuracy of the model. In this applied research, through continuous efforts, the expected goal is basically achieved, and according to the study and research, the design of the original dual-flow model is improved.

• Nuclear Engineering and Technology
• /
• v.54 no.12
• /
• pp.4625-4635
• /
• 2022

A low-energy dielectric loaded accelerator with a non-uniform, multi-segment structure is studied and optimized. So far, no analytical solution is provided for such structures. Also, due to the existing nonlinear behavior and a large number of geometric parameters, the problem of numerical optimizations is complex. For this reason, a method is presented to design and optimize such structures using the Genetic Algorithm (GA). Moreover, the GA output results are compared with Trust Region (TR) and Nelder-Mead Simplex (NMS) methods. Comparative results show that the GA is more efficient in achieving optimization goals and also has a higher speed than the two other methods. Finally, an optimized accelerating tube is integrated into a proper coupler. Then, the accelerator is simulated for full electromagnetic investigations using the CST suite of codes. This design leads to a structure with a power of about 80 kW in the X-band, which delivers electrons to the output energy in the range of 300-459 kV. The length and outer diameter of the accelerating tube obtained are 10 cm and 1 cm, respectively.

• Journal of Korean Neurosurgical Society
• /
• v.61 no.2
• /
• pp.224-232
• /
• 2018

Objective : Cervical ossification of the posterior longitudinal ligament (OPLL) can be treated via anterior or posterior approach, or both. The optimal approach depends on the characteristics of OPLL and cervical curvature. Although most patients can be successfully treated by a single surgery with the proper approach, renewed or newly developed neurological deterioration often requires repeat surgery. Methods : Twenty-seven patients with renewed or newly developed neurological deterioration requiring salvage surgery for multi-segment cervical OPLL were enrolled. Ten patients (group AP) underwent anterior approach, and 17 patients (group PA) underwent posterior approach at the initial surgery. Clinical and radiological data from initial and repeat surgeries were obtained and analyzed retrospectively. Results : The intervals between the initial and repeat surgeries were $102.80{\pm}60.08months$ (group AP) and $61.00{\pm}8.16months$ (group PA) (p<0.05). In group AP, the main OPLL lesions were removed during the initial surgery. There was a tendency that the site of main OPLL lesions causing renewed or newly developed neurological deterioration were different from that of the initial surgery (8/10, p<0.05). Repeat surgery was performed for progressed OPLL lesions at another segment as the main pathology. In group PA, the main OPLL lesions at the initial surgery continued as the main pathology for repeat surgery. Progression of kyphosis in the cervical curvature (Cobb's angle on C2-7 and segmental angle on the main OPLL lesion) was noted between the initial and repeat surgeries. Group PA showed more kyphotic cervical curvature compared to group AP at the time of repeat surgery (p<0.05). Conclusion : The reasons for repeat surgery depend on the type of initial surgery. The main factors leading to repeat surgery are progression of remnant OPLL at a different segment in group AP and kyphotic change of the cervical curvature in group PA.

• Proceedings of the KSPS conference
• /
• 2007.05a
• /
• pp.131-134
• /
• 2007

We propose a multi-stage recognizer architecture that reduces the computation load and makes fast recognizer. To improve performance of baseline multi-stage recognizer, we introduced new feature. We used confidence vector for each phone segment instead of best phoneme sequence. The multi-stage recognizer with new feature has better performance on n-best and has more robustness.

• Journal of Information Processing Systems
• /
• v.14 no.5
• /
• pp.1293-1300
• /
• 2018

For the disadvantages of multi-scale geometric analysis methods such as loss of definition and complex selection of rules in image fusion, an improved multi-focus image fusion method is proposed. First, the initial fused image is quickly obtained based on the lifting stationary wavelet transform, and a simple normalized cut is performed on the initial fused image to obtain different segmented regions. Then, the original image is subjected to NSCT transformation and the absolute value of the high frequency component coefficient in each segmented region is calculated. At last, the region with the largest absolute value is selected as the postfusion region, and the fused multi-focus image is obtained by traversing each segment region. Numerical experiments show that the proposed algorithm can not only simplify the selection of fusion rules, but also overcome loss of definition and has validity.

• Journal of the Korean Society of Radiology
• /
• v.14 no.5
• /
• pp.619-626
• /
• 2020

This study aimed to identify the error rates in Catheter Calibration Mode, Auto Calibration Mode, and Segment Calibration Mode among many calibration modes as a quantitative evaluation tool used for predicting the diameter and length of balloon or stent in percutaneous intravascular balloon dilatation or stent insertion. Our experiment was conducted with Copper Wire of 2 mm × 80 mm (diameter × length) manufactured elaborately for quantitative evaluation in calibration and Metal Ball of 5, 10, 15, 30, and 40 mm and Acryl Phantom of 25 mm, 50 mm, 75mm, 100 mm, 125 mm, 150mm, 175 mm, and 200 mm. At each height, subtraction images were acquired with a cineangiograph and Stenosis Analysis Tool as a software provided by the equipment company was used for measurement. To evaluate the error rates in Catheter Calibration Mode, Copper Wire was put on each acryl phantom before shooting. Copper Wire of 2 mm in diameter was set as a diameter for catheter, and Copper Wire of 8 mm in length was measured with Multi-segments. As a result, the error rates appeared at 1.13 ~ 5.63%. To evaluate the error rates in Auto Calibration Mode, the height of acryl was entered at each height of acryl phantom and the length of 8 mm Copper Wire was measured with Multi-segments and as a result, the error rates appeared at 0 ~ 0.26%. To evaluate the error rates in Segment Calibration Mode, each metal ball on the floor of table was calibrated and the length of 8 mm Copper Wire on each acryl phantom was measured and the length of 8 mm Copper Wire depending on the changes of acryl phantom height was measured with Mutli-segments and as a result, the error rates appeared at 1.05 ~ 19.04%. And in the experiment on OID changes in Auto Calibration Mode, the height of acryl phantom was fixed at 100mm and OID only changed within the range of 450 mm ~ 600 mm and as a result, the error rates appeared at 0.13 ~ 0.38%. In conclusion, it was found that entering the height values in Auto Calibration Mode, among these Calibration Modes for evaluating quantitative vascular dimensions provided by the software was the calibration method with the least error rates and it is thus considered that for calibration using a metal ball or other objects, putting them in the same height as that of treatment sites before calibrating is the method that can reduce the error rates the most.