최근 5년간 약진으로 입원한 환자의 고찰 - 임상양상 및 검사소견(1986∼1990) -

고우석,조광현,윤재일 大韓臨床藥理學會 1993 臨床藥理學會誌 Vol.1 No.1

전파조정기를 이용한 마이크로파 소자용 Calibration

석은영,조현광,서영석,한교용 영남대학교 공업기술연구소 2000 工業技術硏究所論文集 Vol.28 No.1

A microwave device measurement tool kit using the TRL method for microwave tuner was developed. Teflon substrate was used for easy manufacturing. An adapter from CDI was used for the junction transition between microstrip and coaxial line. Mask production procedure which have the 3:1 reduction ratio was developed for microstrip transmission line manufacturing. Using these procedure the Cal Kit which can have Thru, Reflect, and Line was manufactured. TRL error correction method can be applied to the Kit. We tested the Cal kit and the error due to the microstrip line could be corrected. It can be applicable to the accurate measurement of microwave transistors.

인천국제공항 탑승동에 적용 가능한 적정 공조시스템에 관한 연구

석호태,정만석,김지현,여명석,김광우 대한건축학회 2005 대한건축학회논문집 Vol.21 No.1

Generally, indoor thermal environment control consumes much energy in the building Appropriate performance evaluation of the building and HVAC facilities are achieved by an energy simulation The result Which does the energy simulation of property HVAC system in airport Primary, there seems to be reasonable method which save window conduction loads and solar radiation loads on the element for an energy saving Secondly, According to the existence of contact outside in building Loads, the Important factor of influence is divided into solar radiation, window conduction loads and interior loads(lighting, equivalent, occupancy etc) The last, the result which does the energy simulation of each HVAC system, CAV + V AV System is determined profitably in energy performance

흉골골절이 의심되는 외상환자에서 초음파의 진단적 효용성

염석란,김재광,이근,박철완,양혁준,현성열,임용수,진욱 대한응급의학회 2001 대한응급의학회지 Vol.12 No.3

Purpose: This study compared the value of ultrasonography(US) in the diagnosis of sternal fractures with those of conventional radiography and bone scan. Material and Method: From March 2000 to March 2001, in Gil Medical Center, Gachon Medical College, 44 patients(22 women and 22 men, mean age 46 years, range 5∼81 years) complained of tenderness in the sternal area. The cases histories of these patients with clinically suspected sternal fracture due to blunt trauma were prospectively studied. At the time of admission, conventional radiography(sternal view) was done. At one day after admission, sternal ultrasonography was obtained, and after 4∼5 days, a bone scan was done. Sternal ultrasonographic findings, conventional radiographic findings, and bone-scan findings were compared. In addition, a 12-lead ECG was done. When it was necessary, cardiac enzyme was checked and echocardiography was performed. Result: In 21 of 44 patients(48%), a sternal fracture was found by sternal US. Only 9 of 21 patients (43%) were diagnosed as having a sternal fracture by using conventional radiography. 23 of 44 patients (52%) have negative findings on sternal US; they also had negative findings on conventional radiography. Except for 7 unchecked patients, 13 of 21 patients(62%) were diagnosed as having a sternal fracture by bone scan, and among the 23 patients who had negative findings on sternal US, the 19 patients checked with a bone scan also had negative findings. Conclusion: Sternal US can find sternal fractures that are difficult to find on conventional radiography. Compared with bone scan, sternal US takes a shorter time and sternal US is not difficult. Additionally, combined lesions with sternal fractures(peristernal hematoma, soft tissue swelling) and costar cartilage fractures are diagnosed more easily when using sternal US.

還元鐵 粉末을 混合한 카보닐鐵 粉末의 射出成形時 結合劑 選擇

具光德,李奭羲,成章鉉 東亞大學校 大學院 1993 大學院論文集 Vol.18 No.-

To select the optimum binder composition in the metal injection molding of the mixed powders of carbonyl and reduction powders, the PW, CW, SA and HDPE were selected as a multi-component binder and optimum applicable compostion of the multi-component binder was determined during the metal injection molding process. With increasing the HDPE content in the multi-component binders, the moldability(spiral length) decreased. The multi-component binder were condensed separated above 30wt.% of the HDPE. and the improvement of the moldability did not appear above 2wt.% addition of the Sa in the multicomponent binders. Considering the irregular particl size and proous holes at the surface of the reduction powder, it is possible to obtain a sound component for the composition of the binders of 82wt.%PW< 11wt.%CW, 5wT.%HDPE and 2wt.%SA in spited of the secrifice of the st4rength of the green part. The moldability and density decreased with increasing the weight fraction of the reduction powder. The minimum spiral length to obtain a sound green parts of the tensile testing specimen could be determined by using the moldability test. With increasing the weight fraction of the reduction ppoweder, the linear shrinkage decreased and also the sintered density incrased after decreasing at the 10wt.% addition of reduction powders.

실리콘 이온주입 SiO_2층의 나노결정으로 부터의 광루미네센스

김광희,오항석,장태수,권영규,이용현 한국센서학회 2002 센서학회지 Vol.11 No.3

실리콘 기판 위에 형성한 열산화막에 실리콘이온을 주입하고 열처리를 수행한 후, 광루미네센스(photoluminescence:PL) 스펙트럼을 조사하였다. 실리콘 이온도즈의 변화와 열처리 온도의 변화에 따른 PL스펙트럼을 조사하고, 이를 TEM과 XRD 데이터와 비교하여 분석한 결과, 광루미네센스 특성은 산화막내의 실리콘 나노결정으로부터 기인함을 알 수 있었다. 또 산화막을 1분 간격으로 습식 식각하면서 매 식각 시마다 PL스펙트럼을 관측하여 그 변화를 조사하였다. 이러한 실험을 통하여 산화막내에 분포하고 있는 실리콘 나노결정의 크기와 그 수가 PL피크 파장과 강도에 직접적으로 영향을 줌을 알 수 있었다. Photoluminescence(PL) properties of Si^+ -implanted SiO_2 film, which was thermally grown on c-Si substrate, is reported. We have compared room temperature photolurninescence (PL) spectra of the samples which was made in several kinds of implantation, subsequent annealing and SiO_2 film thickness. XRD data was correlated with the PL spectra. Silicon nanocrystals in SiO_2 film is considered as the origin of the photoluminescence. PL spectra was investigated after wet etching of the SiO_2 film by using BOE (Buffered Oxide Etchant) at every one minute. PL peak wavelength was varied as the etching is proceeded. These results indicate that the quantity and the distribution of dominant size of Si nanocrystals in SiO_2 film seem to have a direct effect on PL spectrum.

음성 다이얼링 구성을 위한 기초연구

김현덕,이광석 진주산업대학교 1999 산업과학기술연구소보 Vol.- No.6

We attempt to configuration of voice dialing system, one of speech recognition fields, and verify its performance. The algorithm applied to recognition simulation, is known as dynamic programming method and use TMS320C25 board for signal processing. We applied to 4 digit voice of 10 adults, and verified its performance of voice dialing system.

대화체 음성번역 시스템을 위한 원격 음성입력기 설계

이광석,김현덕 진주산업대학교 산업과학기술연구소 2007 산업과학기술연구소보 Vol.- No.14

음성 인식기를 사용할 때에 마이크의 위치에 구애받지 않고, 자동적으로 음성을 검출하는 음성 입력 방식은 사용자의 측면에서 대단히 편리하다. 본 연구에서는 대화체 음성 번역 시스템에 이와 같은 원격음성 입력 기술을 구현하기 위하여 자동 음성부 검출 방법과 마이크 어레이를 이용한 다채널 잡음 감소 방법을 적용하였다. 신호 대 잡음비가 25[dB]인 상황에서, 음성부의 시작점과 끝점이 기준점의 200[msec] 이내로 검출될 정확도는 약 97%이었다. 지연-가산 방법을 이용한8 채널 마이크 어레이 잡음 감소기는 최대 6[dB]의 신호 대 잡음비의 개선을 나타내었으며 음성 인식에서의 오류 감축률은 약 16% 정도를 보였다.

PRMA 성능분석을 위한 해석모델

이광석,김현덕 진주산업대학교 2000 論文集 Vol.39 No.-

This paper presents a analytic model for performance test of PRMA(Packet Reservation Multiple Access) protocol as speech-text service. Speech as input traffic is assumed as on-off model exponential distribution in talkspurt and silence duration, and is assumed data packets as rising event in constant probability. The main parameter for system performatnce test concerned voice packet dropping probability, data packet delay and system throughput of data packet.

잡음환경에서의 음성인식을 위한 스펙트럴 가중치 부과 방식

이광석,김현석 진주산업대학교 2000 산업과학기술연구소보 Vol.- No.7

This paper is concerned for improving the result of speech recognition under the noisy speech. We knew that spectral subtraction and recovery of valleys in spectral envelepe obtained from noisy speech are more effective for the improvement of the recognition. In this paper, the averaged spectral envelope obtained from vowel spectrums are used for the emphasis of valleys. The vocalic spectral information at lower frequency range is emphasized and the spectrum obtained from consonants is not changed. In experiment, the emphasis coefficients are varied on cepstral domain. This method is used for the recognition of noisy digits and is improved.