• Journal of the Korean Society of Radiology
• /
• v.14 no.6
• /
• pp.773-780
• /
• 2020

With the development of the 4th industrial, research is being conducted to prevent diseases and reduce damage in various fields of science and technology such as medicine, health, and bio. As a result, artificial intelligence technology has been introduced and researched for image analysis of radiological examinations. In this paper, we will directly apply a deep learning model for classification and detection of pneumonia using chest X-ray images, and evaluate whether the deep learning model of the Inception series is a useful model for detecting pneumonia. As the experimental material, a chest X-ray image data set provided and shared free of charge by Kaggle was used, and out of the total 3,470 chest X-ray image data, it was classified into 1,870 training data sets, 1,100 validation data sets, and 500 test data sets. I did. As a result of the experiment, the result of metric evaluation of the Inception V3 deep learning model was 94.80% for accuracy, 97.24% for precision, 94.00% for recall, and 95.59 for F1 score. In addition, the accuracy of the final epoch for Inception V3 deep learning modeling was 94.91% for learning modeling and 89.68% for verification modeling for pneumonia detection and classification of chest X-ray images. For the evaluation of the loss function value, the learning modeling was 1.127% and the validation modeling was 4.603%. As a result, it was evaluated that the Inception V3 deep learning model is a very excellent deep learning model in extracting and classifying features of chest image data, and its learning state is also very good. As a result of matrix accuracy evaluation for test modeling, the accuracy of 96% for normal chest X-ray image data and 97% for pneumonia chest X-ray image data was proven. The deep learning model of the Inception series is considered to be a useful deep learning model for classification of chest diseases, and it is expected that it can also play an auxiliary role of human resources, so it is considered that it will be a solution to the problem of insufficient medical personnel. In the future, this study is expected to be presented as basic data for similar studies in the case of similar studies on the diagnosis of pneumonia using deep learning.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.54 no.1
• /
• pp.96-110
• /
• 2017

Most of vehicle detection studies using the existing general lens or wide-angle lens have a blind spot in the rear detection situation, the image is vulnerable to noise and a variety of external environments. In this paper, we propose a method that is detection in harsh external environment with noise, blind spots, etc. First, using a fish-eye lens will help minimize blind spots compared to the wide-angle lens. When angle of the lens is growing because nonlinear radial distortion also increase, calibration was used after initializing and optimizing the distortion constant in order to ensure accuracy. In addition, the original image was analyzed along with calibration to remove fog and calibrate brightness and thereby enable detection even when visibility is obstructed due to light and dark adaptations from foggy situations or sudden changes in illumination. Fog removal generally takes a considerably significant amount of time to calculate. Thus in order to reduce the calculation time, remove the fog used the major fog removal algorithm Dark Channel Prior. While Gamma Correction was used to calibrate brightness, a brightness and contrast evaluation was conducted on the image in order to determine the Gamma Value needed for correction. The evaluation used only a part instead of the entirety of the image in order to reduce the time allotted to calculation. When the brightness and contrast values were calculated, those values were used to decided Gamma value and to correct the entire image. The brightness correction and fog removal were processed in parallel, and the images were registered as a single image to minimize the calculation time needed for all the processes. Then the feature extraction method HOG was used to detect the vehicle in the corrected image. As a result, it took 0.064 seconds per frame to detect the vehicle using image correction as proposed herein, which showed a 7.5% improvement in detection rate compared to the existing vehicle detection method.

• Nuclear Medicine and Molecular Imaging
• /
• v.43 no.5
• /
• pp.468-477
• /
• 2009

Purpose: Hydrodynamic-based procedure is a simple and effective gene delivery method to lead a high gene expression in liver tissue. Non-invasive imaging reporter gene system has been used widely with herpes simplex virus type 1 thymidine kinase (HSV1-tk) and its various substrates. In the present study, we investigated to image the expression of HSV1-tk gene with 5-(2-iodovinyD-2'-deoxyuridine (IVDU) in mouse liver by the hydrodynamicbased procedure. Materials and Methods: HSV1-tk or enhanced green fluorescence protein (EGFP) encoded plasmid DNA was transferred into the mouse liver by hydrodynaminc injection. At 24 h post-injection, RT-PCR, biodistribution, fluorescence imaging, nuclear imaging and digital wholebody autoradiography (DWBA) were performed to confirm transferred gene expression. Results: In RT-PCR assay using mRNA from the mouse liver, specific bands of HSV1-tk and EGFP gene were observed in HSV1-tk and EGFP expressing plasmid injected mouse, respectively. Higher uptake of radiolabeled IVDU was exhibited in liver of HSV1-tk gene transferred mouse by biodistribution study. In fluorescence imaging, the liver showed specific fluorescence signal in EGFP gene transferred mouse. Gamma-camera image and DWBA results showed that radiolabeled IVDU was accumulated in the liver of HSV1-tk gene transferred mouse. Conclusion: In this study, hydrodynamic-based procedure was effective in liver-specific gene delivery and it could be quantified with molecular imaging methods. Therefore, co-expression of HSV1-tk reporter gene and target gene by hydrodynamic-based procedure is expected to be a useful method for the evaluation of the target gene expression level with radiolabeled IVDU.

• 한국지구과학회:학술대회논문집
• /
• 2010.04a
• /
• pp.88-89
• /
• 2010

최근 인간의 활동범위와 영역이 확대되고 산업이 발전하면서 인간의 삶과 지속가능한 발전 등 도시 기후에 관한 관심도 높아지고 있다. 산업혁명 이후 도시화와 산업화로 인해 인구가 증가하고 도시지역으로 집중됨으로써 도시 열섬화 현상에 대한 도시환경문제가 부각되고 있다. 이는 최근까지도 도시개발에 있어서 기능과 효율성이 우선시 되어 도시기후에 대한 배려가 이루어지지 못하고 있으며, 오히려 과도한 냉난방을 가동하는 등 쾌적한 실내 환경 조성을 위한 노력만을 행해왔다. 도시화에 따른 도시의 열환경 구조의 변화는 토지이용의 변화에 따른 피복상태와 밀접한 관련이 있다는 연구들이 수행된 바 있다. 이렇듯 도시화가 진행됨에 따라서 도심 지표면을 덮고 있는 포장재도 변하고 있다. 대표적인 토지피복재로는 콘크리트와 아스팔트 등의 인공포장재, 수계, 삼림 등으로 크게 나누어 볼 수 있다. 최근 도심의 발달로 인해 도심의 표면은 점차 인공포장재인 아스팔트와 콘크리트로 덮여지고 있다. 인공포장재는 맑은 여름철 낮에 받아들인 열을 야간에도 머금고 있어 도시열섬현상의 주요원인이 된다. 도시화가 진행됨에 따라 토지이용형태가 변화하고 있으며 이러한 토지피복의 변화는 그 지역의 기온과 풍향, 풍속뿐만 아니라 지표온도도 변화시키므로 도시 열환경 구조에 적지 않은 영향을 미치고 있다. 과거에는 자연 환경과 도시공간에 대한 인식이 다른 분야로 나누어져서 다루었지만 현재 위성영상 기술의 발달로 많은 공간 정보를 파악할 수 있게 된 바 도시기후변화에 더욱 직접적이고 근본적인 접근이 쉬워졌다. 원격탐사기법의 활용은 위성자료를 이용하여 동시간대 평면적인 열구조를 정량적으로 파악하는데에 중요한 자료를 제공하여 도시지역을 덮고 있는 인공자재의 존재가 도시열섬의 형성과 밀접하게 연관이 있다는 사실을 짐작할 수 있다. 따라서 도시기후변화의 문제점을 더욱 적극적으로 해결하기 위해서는 토지이용에 따른 지표면 온도 상승의 현황을 파악하고 이를 저감 시킬 수 있는 대책들이 수립되어야 한다. 본 연구는 보다 세분화된 도시 열환경을 정량적으로 분석 평가하기 위해서 토지피복별 분류를 3가지로 대구시 중구 경북대학교 부속 고등학교(이하 사대부고 지점)를 도심지역으로, 경상남도 창녕군 창녕읍 우포늪(이하 우포지점)을 수계지점으로, 경상북도 안동시 길안면 만음리(이하 안동지점) 지점과 대구시 칠곡군 동명면 득명리 팔공산 한티재 도립공원(이하 팔공지점)을 산림으로 분류하여 연구하였다. 대구 계명대학교 기후환경연구실에서 보유하고 있는 AWS(Automatic Weather Station) 자료로 기상요소를 분석하였으며, MODIS Terra 위성영상을 이용하여 지표온도를 추출하고 분석하였다. 또 기상요소와 지표온도를 이용해 회귀식을 도출하여 추정기온을 산출하였다. 그 결과 첫째, 계절에 따른 기온의 시간변화는 여름의 평균기온이 $25.13^{\circ}C$와 $24.12^{\circ}C$로 사대지점과 우포지점의 평균기온이 가장 높게 나타났으며, 이는 도심에서 발생되는 인공열의 영향으로, 우포지점은 수계의 특징이 반영된 결과라 할 수 있다. 둘째, 계절에 따른 풍속의 시간변화는 여름의 경우 우포지점의 풍속이 1.63m/s로 가장 높은 반면 안동지점의 풍속이 0.27m/s로 가장 낮은 것으로 나타났다. 겨울의 경우 팔공지점의 풍속이 1.82m/s로 가장 높게 나타났다. 토지피복에 따른 지표면의 변화가 도시기후에 미치는 영향을 정량적으로 평가하고, 또 지표면 온도와 기온과의 차이를 알아보기 위하여 MODIS 위성 영상을 이용하여 세 지점을 대상으로 토지피복에 따른 열환경을 평가 분석하여 다음과 같은 결론을 얻을 수 있었다. 첫째, MODIS 위성영상을 이용하여 산출한 지표면 온도는 여름철 주간에 안동지점의 경우 주변지역에 비해 지표면 온도가 약 $26^{\circ}C$로 낮게 나타났으며 우포지점의 경우 수계가 가지는 열 완충능력으로 약 $27^{\circ}C$의 낮은 지표면 온도를 나타내었다. 사대지점의 경우 약 $34^{\circ}C$이상의 높은 지표면 온도를 나타내었다. 둘째, MODIS 위성영상을 이용하여 산출한 지표면 온도와 관측된 기온과의 회귀식을 도출하여 상관분석 한 결과, 모든 지점의 값에서 상관성 및 신뢰도가 높은 것으로 나타났다. 셋째, 상관분석의 결과를 통하여 추정한 기온은 지표면 온도와의 차이가 있지만 유사한 패턴의 결과로 추출되었다. 이러한 결과로 볼 때 도시의 인공자재를 이용한 건축과 개발이 도시열섬현상을 유발하는데 중요한 역할을 하는 것을 정량적으로 평가할 수 있었다. 따라서 본 논문의 연구결과를 바탕으로 도시계획에 있어서 인공구조물에 의한 기온과 풍속이 받는 영향을 고려하여 도심의 인공구조물의 배치나 자재에 대한 개발이 이루어져야 할 것이며 열교환의 방해 및 바람순환이 확보되는 구조로 개선되어야 할 것이다.

• Investigative Magnetic Resonance Imaging
• /
• v.9 no.1
• /
• pp.24-29
• /
• 2005

Purpose : Instead of conventional two-dimensional (2-D) visual stimuli, three-dimensional (3-D) visual stimuli with stereoscopic vision were employed for the study of functional Magnetic Resonance Imaging (f-MRI). In this paper f-MRI with 3-D visual stimuli is investigated in comparison with f-MRI with 2-D visual stimuli. Materials and Methods : The anaglyph which generates stereoscopic vision by viewing color coded images with red-blue glasses is used for 3-D visual stimuli. Two-dimensional visual stimuli are also used for comparison. For healthy volunteers, f-MRI experiments were performed with 2-D and 3-D visual stimuli at 3.0 Tesla MRI system. Results : Occipital lobes were activated by the 3-D visual stimuli similarly as in the f-MRI with the conventional 2-D visual stimuli. The activated regions by the 3-D visual stimuli were, however, larger than those by the 2-D visual stimuli by $18\%$. Conclusion : Stereoscopic vision is the basis of the three-dimensional human perception. In this paper 3-D visual stimuli were applied using the anaglyph. Functional MRI was performed with 2-D and 3-D visual stimuli at 3.0 Tesla whole body MRI system. The occipital lobes activated by the 3-D visual stimuli appeared larger than those by the 2-D visual stimuli by about $18\%$. This is due to the more complex character of the 3-D human vision compared to 2-D vision. The f-MRI with 3-D visual stimuli may be useful in various fields using 3-D human vision such as virtual reality, 3-D display, and 3-D multimedia contents.

• The Transactions of the Korea Information Processing Society
• /
• v.2 no.6
• /
• pp.937-946
• /
• 1995

We proposed and efficient block segmentation and classification method for the document analysis using SGLDM(spatial gray level dependence matrix) and BP (back Propagation) neural network. Seven texture features are extracted directly from the SGLDM of each gray-level block image, and by using the nonlinear classifier of neural network BP, we can classify document blocks into 9 categories. The proposed method classifies the equation block, the table block and the flow chart block, which are mostly composed of the characters, out of the blocks that are conventionally classified as non-character blocks. By applying Sobel operator on the gray-level document image beforebinarization, we can reduce the effect of the background noises, and by using the additional horizontal-vertical smoothing as well as the vertical-horizontal smoothing of images, we can obtain an effective block segmentation that does not lead to the segmentation into small pieces. The result of experiment shows that a document can be segmented and classified into the character blocks of large fonts, small fonts, the character recognigible candidates of tables, flow charts, equations, and the non-character blocks of photos, figures, and graphs.

• Korean Journal of Remote Sensing
• /
• v.30 no.6
• /
• pp.829-837
• /
• 2014

The KOMPSAT-3 can acquire panchromatic stereo image with 0.7 m spatial resolution, and provides Rational Polynomial Coefficient (RPC). In order to determine ground coordinate using the provides RPC, which include interior-exterior orientation errors, its adjustment is needed by using the Ground Control Point (GCP). Several thousands of national Unified Control Points (UCPs) are established and overall distributed in the country by the Korean National Geographic Information Institute (NGII). UCPs therefore can be easily searched and downloaded by the national-control-point-record-issues system. This paper introduced the point-extraction method and the distance-bearing method to detect of UCPs. As results, the distance-bearing method was better detected through the experiment. RPC adjustment using this method was compared with that by only one UCP and GCPs using GPS. The proposed method was more accurate than the other method in the horizontal. As demonstrated in this paper, the proposed UCPs detection method could be replaced GPS surveying for RPC adjustment.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.44 no.1
• /
• pp.10-16
• /
• 2007

In this paper, we propose a technique for retrieving images using spatial color correlation and texture characteristics based on local fourier transform. In order to retrieve images, two new descriptors are proposed. One is a color descriptor which represents spatial color correlation. The other is a descriptor combining the proposed color descriptor with texture descriptor. Since most of existing color descriptors including color correlogram which represent spatial color correlation considered just color distribution between neighborhood pixels, the structural information of neighborhood pixels is not considered. Therefore, a novel color descriptor which simultaneously represents spatial color distribution and structural information is proposed. The proposed color descriptor represents color distribution of Min-Max color pairs calculating color distance between center pixel and neighborhood pixels in a block with 3x3 size. Also, the structural information which indicates directional difference between minimum color and maximum color is simultaneously considered. Then new color descriptor(min-max color correlation descriptor, MMCCD) containing mean and variance values of each directional difference is generated. While the proposed color descriptor includes by far smaller feature vector over color correlogram, the proposed color descriptor improves 2.5 % ${\sim}$ 13.21% precision rate, compared with color correlogram. In addition, we propose a another descriptor which combines the proposed color descriptor and texture characteristics based on local fourier transform. The combined method reduces size of feature vector as well as shows improved results over existing methods.

• The KIPS Transactions:PartB
• /
• v.14B no.4
• /
• pp.249-254
• /
• 2007

A mobile robot with various types of sensors and wireless camera is introduced. We show this mobile robot can detect objects well by combining the results of active sensors and image processing algorithm. First, to detect objects, active sensors such as infrared rays sensors and supersonic waves sensors are employed together and calculates the distance in real time between the object and the robot using sensor's output. The difference between the measured value and calculated value is less than 5%. We focus on how to detect a object region well using image processing algorithm because it gives robots the ability of working for human. This paper suggests effective visual detecting system for moving objects with specified color and motion information. The proposed method includes the object extraction and definition process which uses color transformation and AWUPC computation to decide the existence of moving object. Shape information and signature algorithm are used to segment the objects from background regardless of shape changes. We add weighing values to each results from sensors and the camera. Final results are combined to only one value which represents the probability of an object in the limited distance. Sensor fusion technique improves the detection rate at least 7% higher than the technique using individual sensor.

• Cartoon and Animation Studies
• /
• s.43
• /
• pp.123-150
• /
• 2016

There has been discussed very little about animation authorship studies, especially the themes of authors and their original animation texts since 1990s. This study is to explore the academic discourse of animation authorship studies understanding the media aesthetics of animation and the scholarly approach of animation studies about authorship. This article examines 55 articles via meta analysis about animation authorship studies drawn from the 1,516 articles on the general animation studies. The results were as follows. First, the domestic animation studies on the authorship were made about animators of Japan, US, and european countries. Second, It was dominant that scholarly interest on the screen direction and visual expression of animation texts. This shows the authorship approaches were mainly about visual aspects of animations. The domestic animation authorship studies did not trace the authorship issues on author's world view and visual style revealed in the corpus of texts. Instead, they discussed authorship issues on the characteristics of individual animation text. It has been evident that animation studies were not enthusiastic about building the independent theory on animation. Therefore, animation studies have tried theorizing the animation issues borrowing the literature and film theories. This study can contribute to increase phase of animation studies by drawing the intensive discussion of animation authors and their authorship.