• Journal of KIISE:Software and Applications
• /
• v.28 no.3
• /
• pp.285-297
• /
• 2001

본 논문에서는 특정한 객체의 색 분포 모델링으로부터 얻어지는 가중 색 히스토그램과 지배적인 색의 영상공간 분포특성을 이용한 내용기반 영상 검색 방법을 제안한다. 특정한 객체의 예로 사람 얼굴을 선택했고, 그것의 색 분포를 u*-v* 색도 공간에서 모델링 했으며, 모델의 정규화된 부피를 균등 양자화된 색도 공간의 각 빈(bin)의 히스토그램 값에 대한 가중치로 결정하고, 결정된 가중치를 히스토그램 정합 과정에 적용하였다. 또한 색 히스토그램 값이 큰 특정한 수의 빈으로 정의되는 지배적인 색의 영상 공간 분포를 가중 색 히스토그램과 함께 유사성의 측정기준으로 사용하였다. 제안한 검색 방법을 500여개의 영상에 대해 실험한 결과 제안한 방법이 얼굴을 포함하는 영상을 질의로 주었을 때 얼굴을 포함하는 영상을 우선적으로 찾는데 효과적임을 확인하였다.

• Journal of the Korean Society of Radiology
• /
• v.11 no.5
• /
• pp.379-385
• /
• 2017

This study identifies the optimal tube voltages depending on the changes in the patient's body type for limb tests using a digital radiography (DR) system. For the upper-limp test, the dose area product (DAP) was fixed at $5.06dGy{\ast} cm^2$, and for the lower-limb test, the DAP was fixed at $5.04dGy{\ast} cm^2$. Afterwards, the tube voltage was changed to four different stages and the images were taken three times at each stage. The thickness of the limbs was increased by 10 mm to 30 mm to change in the patient's body type. For a quantitative evaluation, Image J was used to calculate the contrast to noise ratio (CNR) and signal to noise ratio (SNR) among the four groups, according to the tube voltage. For statistical testing, the statistically significant differences were analyzed through the Kruskal-Wallis test at a 95% confidence level. For the qualitative analysis of the images, the pre-determined items were evaluated based on a 5-point Likert scale. In both upper-limb and lower-limb tests, the more the tube voltage increased, the more the CNR and SNR of the images decreased. The test on the changes depending on the patient's body shape showed that the more the thickness increased, the more the CNR and SNR decreased. In the qualitative evaluation on the upper limbs, the more the tube voltage increased, the more score increased to 4.6 at the maximum of 55kV and 3.6 at 40kV, respectively. The mean score for the lower limbs was 4.4, regardless of the tube voltage. The more either the upper or lower limbs got thicker, the more the score generally decreased. The score of the upper limps sharply dropped at 40kV, whereas that of the lower limps sharply dropped at 50kV. For patients with a standard thickness, the optimized images can be obtained when taken at 45kV for the upper limbs, and at 50kV for the lower limbs. However, when the thickness of the patient's limbs increases, it is best to set the tube voltage at 50 kV for the upper limbs and at 55 kV for the lower limbs.

• Journal of Korean Society of Transportation
• /
• v.30 no.2
• /
• pp.127-136
• /
• 2012

Multi-car rear-end collision accidents have three categories: sequential collision from the rear which is commonly referred to as chain reaction collision, sequential collision from the front, and mixed-order collision. This paper suggests several effective methods of reconstruction analysis for multi-car rear-end collision accidents. First, by incorporating the traditional empirical method which uses vehicle damage caused by brake dive and passenger injuries, with results of theoretical analysis made within mechanics of rigid body, it is made possible for the method to be put to immediate practical use. A methodology to precisely analyze multi-car rear-end collision accidents was suggested using a simulation program simultaneously with a video event data recorder which is starting to be widely used in domestic vehicles. To go beyond the simple intuitive analysis of the video event data recorder, the simulation analysis based on the results of video analysis was executed to acquire various information, so that the causes and responsibility could be clearly stated.

• Proceedings of the KIEE Conference
• /
• 2002.11b
• /
• pp.177-179
• /
• 2002

본 논문에서는 비접지 계통에서 주로 사용되는 선택 지락 계전기를 디지털화 함으로써 구현할 수 있는 성능 향상 방법을 제시한다. 선택 지락 계전기는 영상 전압과 영상 전류를 이용하여 지락 사고를 검출하지만, 지락 사고시 계전점의 역상 전류가 영상 전류보다 큰 값을 가지며 위상도 앞서게 되므로, 역상 전류를 사용하는 것이 바람직하다. 6.6 kV 비접지 계통에서의 지락 사고를 EMTP를 이용하여 모의하였으며. 1회선 계통과 2회선 계통에 대한 계전점의 영상 전류와 역상 전류를 비교하여 이를 검증하였다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.20 no.6
• /
• pp.1183-1188
• /
• 2016

Kinect sensor version 2 is a kind of camera released by Microsoft as a computer vision and a natural user interface for game consoles like Xbox one. It allows acquiring color images, depth images, audio input and skeletal data with a high frame rate. In this paper, using depth image, we present a surveillance system of a certain area within Kinect's field of view. With computer vision library(Emgu CV), if an object is detected in the target area, it is tracked and kinect camera takes RGB image to send it in database server. Therefore, a mobile application on android platform was developed in order to notify the user that Kinect has sensed strange motion in the target region and display the RGB image of the scene. User gets the notification in real-time to react in the best way in the case of valuable things in monitored area or other cases related to a reserved zone.

• Investigative Magnetic Resonance Imaging
• /
• v.14 no.1
• /
• pp.56-63
• /
• 2010

Purpose : DIR image is relatively free from susceptibility artifacts therefore, DIR image can make it possible to reliably measure cortical thickness/volume. One drawback of the DIR acquisition is the long scan time to acquire the fully sampled 3D data set. To solve this problem, we applied a parallel imaging method (GRAPPA) and verify the reliability of using the volumetric study. Materials and methods : Six healthy volunteers (3 males and 3 females; age $25.33{\pm}2.25$ years) underwent MRI using the 3D DIR sequence at a 3.0T Siemens Tim Trio MRI scanner. GRAPPA simulation was performed from the fully sampled data set for reduction factor 2. Data reconstruction was performed using MATLAB R2009b. Freesurfer v.4.3.0 was used to evaluate the cortical thickness of the entire brain, and to extract white matter information from the DIR image, Analyze 9.0 was used. The global cortical thickness estimated from the reconstructed image was compared with reference image by using a T-test in SPSS. Results : Although reduced SNR and blurring are observed from the reconstructed image, in terms of segmentation the effect was not so significant. The volumetric result was validated that there were no significant differences in many cortical regions. Conclusion : This study was performed with DIR image for a volumetric MRI study. To solve the long scan time of 3D DIR imaging, we applied GRAPPA algorithm. According to the results, fast imaging can be done with reduction factor 2 with little loss of image quality at 3.0T.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.13 no.5
• /
• pp.313-322
• /
• 2020

A large volume of data will improve the robustness of deep learning models and avoid overfitting problems. In automatic tongue segmentation, the availability of annotated tongue images is often limited because of the difficulty of collecting and labeling the tongue image datasets in reality. Data augmentation can expand the training dataset and increase the diversity of training data by using label-preserving transformations without collecting new data. In this paper, augmented tongue image datasets were developed using seven augmentation techniques such as image cropping, rotation, flipping, color transformations. Performance of the data augmentation techniques were studied using state-of-the-art transfer learning models, for instance, InceptionV3, EfficientNet, ResNet, DenseNet and etc. Our results show that geometric transformations can lead to more performance gains than color transformations and the segmentation accuracy can be increased by 5% to 20% compared with no augmentation. Furthermore, a random linear combination of geometric and color transformations augmentation dataset gives the superior segmentation performance than all other datasets and results in a better accuracy of 94.98% with InceptionV3 models.

• Progress in Medical Physics
• /
• v.21 no.1
• /
• pp.93-98
• /
• 2010

The aim of this work was to establish the methodology for event positioning by measuring depth of interaction (DOI) information and to evaluate the system sensitivity and spatial resolution of the new detector for I-125 and Tc-99m imaging. For this purpose, a Monte Carlo simulation tool, DETECT2000 and GATE were used to model the energy deposition and light distribution in the detector and to validate this approach. Our proposed detector module consists of a monolithic CsI(Tl) crystal with dimensions of $50.0{\times}50.0{\times}3.0\;mm^3$. The results of simulation demonstrated that the resolution is less than 1.5 mm for both I-125 and Tc-99m. The main advantage of the proposed detector module is that by using 3 mm thick CsI(Tl) with maximum-likelihood position-estimation (MLPE) method, high resolution I-125 imaging and high sensitivity Tc-99m imaging are possible. In this paper, we proved that our new detector to be a reliable design as a detector for a multi-energy SPECT.

• The Journal of the Acoustical Society of Korea
• /
• v.35 no.3
• /
• pp.175-182
• /
• 2016

For application to several MHz photoacoustic imaging systems, a needle hydrophone was designed and fabricated by using PMN-PZT piezoelectric single crystal, and its characteristics were evaluated through comparison with a commercial PVDF(Polybinylidene Fluoride) hydrophone of which receiving sensitivity is known. The simulation using the KLM model results show that the peak receiving impulse response for $50{\Omega}$ terminating impedance of the fabricated hydrophone is -261.6 dB re $1V/{\mu}Pa$ and the frequency response is relatively flat over 2 ~ 12 MHz with fluctuation less than 5 dB. The measurement results using tone burst signals also show that it has higher (ave. 10.9 dB) sensitivity than the commercial hydrophone in 2 ~ 8 MHz, and the receiving sensitivity of $-255.8{\pm}2.8$ dB re $1V/{\mu}Pa$ was measured for the fabricated hydrophone. In addition, it is known that the photoacoustic signals and the image of a hair obtained by a mechanical scanned photoacoustic imaging system with the fabricated hydrophone were bigger and better than those obtained with the commercial hydrophone.

• Journal of the Korean Society of Radiology
• /
• v.12 no.3
• /
• pp.289-295
• /
• 2018

Most diagnostic devices in the medical field use X-ray sources, which emit energy spectra. In radiological diagnosis, the quantitative and qualitative analyses of X-rays are essential for maintaining the image quality and minimizing the radiation dose to patients. This work aims to obtain the X-ray energy spectra used in diagnostic imaging by Monte Carlo simulation. Various X-ray spectra are simulated using a Monte Carlo simulation tool. These spectra are then compared to the reference data obtained with a tungsten anode spectral model using the interpolating polynomial (TASMIP) code. The X-ray tube voltages used are 50, 60, 80, 100, and 110 kV, respectively. CdTe and a-Se detector are used as the detectors for obtaining the X-ray spectra. Simulation results demonstrate that the various X-ray spectra are well matched with the reference data. Based on the simulation results, an appropriate X-ray spectrum, in accordance with the tube voltage, can be selected when generating an image for diagnostic imaging. The dose to be delivered to the patient can be predicted prior to examination in the diagnostic field.