• 동굴
• /
• 제68호
• /
• pp.75-98
• /
• 2005

Korea physiographic province is divided into two provinces which is northern Chugaryung graben zone and southern Chugraryung graben zone. Northern Chugraryung is also divided in to Gema block and Kohan block, and southern Chugraryung dividedinto Han block, Yongnam basin and Honam plain. The above mentioned macro geomorophic units is divided, mainly on the geotectonics. The meso geomorphic units is divided, based upon the regional distribution of topographic characteristic that is plateaus, mountains, mountain range, basins and great plain etc. Micro geomorphic units id into a mountain, a hill, a plain, and a lowland, and then it is formed by selfreliant topographic unit. And micro topographic (fan, peneplain, delta, etc) dealt with a characteristics unit. In this article has a disregarded amallest scale that is included flood plait natural levee, back marsh and oxbow lake etc. Accordingly, it shows macro units are meso units are 5, meso units are 53, micro units are 299. A study method of physiographic provincs prefered to aufsteigende and abstergands methoy. How to organically combine topographic factors can be seen in regional distribution of the peculiar topographic characteristics, for charage teristic of topographic makes a study on the topographic of micro unit such understanding as aufsteigende method. At the same time, since it can be studied systematically from marco unit to micro unit like the absteigende methods, I used both methods. And this establishment of physiographic province based on the scientific method depend on the base map of climate classification. Geology, Soil, Biology. I feel confident that it will be used the basic map for land use map, land classification map, study of geomorphology of Korea. And will be used for study of a topographic standard data.

• 한국통신학회논문지
• /
• 제36권11C호
• /
• pp.662-672
• /
• 2011

본 논문은 새로운 경계선 보존 알고리즘을 이용하여 블록화 현상을 제거하는 디블로킹 필터와 HD해상도의 실시간 영상처리가 가능한 디블로킹 필터의 VLSI구조를 제안한다. 기존의 블록 분류 기반의 접근 방법과 달리 제안된 알고리즘은 픽셀 분류 기반 접근을 사용한다. 또한 제안된 경계선 보존 맵은 픽셀을 경계선 영역과 평탄 영역으로 분류하며, 블록화 현상 제거에 사용되는 오프셋 필터와 경계선 보존 필터의 기반이 된다. 이를 바탕으로 제안된 디블로킹 필터의 VLSI구조는 고연산량 처리를 위하여 블록 전체에 파이프라인 기법을 적용하였다. 또한 블록 버퍼를 위한 메모리 절감 구조는 메모리의 사용을 최적화 시킨다. 본 필터는 VHDL을 이용한 설계를 통하여 CycloneII FPGA상에서 구현된 구조의 동작을 검증 후, Synopsys의 Design Compiler와 ANAM 0.25 ${\mu}m$ CMOS cell library로 합성하여 칩으로 구현하였을 때의 성능을 예측하였다. 제안된 알고리즘의 실험 결과는 세밀한 영상성분을 보존하면서 효과적으로 블록화 현상을 제거하며, 픽셀 분류 기반에서 제안된 알고리즘은 블록 분류 기반보다 PSNR 성능이 우수함을 보였다.

• 한국통신학회논문지
• /
• 제36권12B호
• /
• pp.1574-1585
• /
• 2011

본 논문에서 지능형 안전 자동차 시스템을 위해 연산량를 줄인 속도 제한 표지판 실시간 인식 방법을 제안한다. 제안된 방법은 관심영역의 전체 픽셀 정보를 특징으로 사용한 기존 방법의 큰 연산량을 줄이기 위해 적은 수의 DCT 계수를 선택하고, 격자구조로 분할된 영상에 대해 Block-DCT를 이용하여 산술 연산을 효과적으로 줄였다. 제안된 알고리듬은 연산량을 줄이기 위해 제안된 상관계수와 분산을 이용한 판별식에 따라 DCT 계수를 선택하고 이를 선형 판별법과 Mahalanobis Distance를 이용하여 속도 제한 표지판을 인식한다. 인식 성능을 높이기 위해 연속 프레임의 누적 분류 결과를 사용한다. 실험 결과 연속된 프레임에 대하여 100.0 %의 인식률을 보이며 기존 방식 대비 곱셈 연산량은 69.3 %, 덧셈은 67.9 % 감소를 확인할 수 있었다.

• 반도체공학회 논문지
• /
• 제2권1호
• /
• pp.17-20
• /
• 2024

본 논문은 전통적인 자원 집약적인 컴퓨터 비전 모델의 한계를 해결하기 위해 저전력 엣지 장치에 최적화된 새로운 경량 객체 검출 모델을 제안합니다. 제안된 검출기는 Single Shot Detector (SSD)에 기반하여 소형이면서도 견고한 네트워크를 설계하였고, 작은 객체를 효율적으로 감지하는 데 있어 효율성을 크게 향상시키도록 모델을 구성하였다. 이 모델은 주로 두 가지 구성요소로 구성되어 있습니다: Depthwise 와 Pointwise Convolution 레이어를 사용하여 효율적인 특징 추출을 위한 Light_Block, 그리고 작은 객체의 향상된 감지를 위한 Enhancer_Block 으로 나누었다. 우리의 모델은 300x480 의 이미지 크기를 가진 Udacity 주석이 달린 데이터셋에서 처음부터 훈련되었으며, 사전 훈련된 분류 가중치의 필요성을 제거하였다. 약 0.43M 의 파라미터로 5.5MB 만의 무게를 가진 우리의 검출기는 평균 정밀도 (mAP) 27.7%와 140 FPS 의 처리 속도를 달성하여, 정밀도와 효율성 모두에서 기존 모델을 능가하였다. 따라서, 본 논문은 추론의 정확성을 손상시키지 않으면서 엣지 장치를 위한 객체 검출에서의 효과적인 경량화를 보여주고 있다.

• The Korean Journal of Pain
• /
• 제5권2호
• /
• pp.273-278
• /
• 1992

With the medical progress that has given spinal cord injured(SCI) individuals greater longevity and better overall health, chronic pain is emerged as major challenge in treating this population. According to past reports, estimates of prevalance of severe/disabling chronic pain in SCI patients have ranged from 18% to 63%. In etiologies of chronic pain in SCI patients, psychic or psychogenic pain categories should be included and more recent data have demonstrated that the persistant pain is directly related to higher levels of psychosocial distress and impairment. Recently, neurophysiological classification of the SCI pain syndrome into three etiologic groups(a; mechanical pain, b; radicular pain, c; deafferentation pain) is more frequently adopted for the classification of chronic SCI pain syndrome. The deafferentation pain is most common of the pain syndromes associated with SCI. After cervical epidural anesthesia for the surgical intervention of decubitus ulcer on the hip of two SCI patients, there were much reduction of existing chronic deafferentation character pain.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2008년도 하계종합학술대회
• /
• pp.903-904
• /
• 2008

This paper suggests a simple scheme of block motion estimation in which the search pattern selection is based on the classification of motion content available in the spatio temporal neighboring blocks. The search area is divided into eight sectors and the search pattern selection is also based on the direction of predicted motion vector. Experimental results show that the proposed algorithm has achieved good predicted image quality measured in terms of PSNR and has very less computational complexity.

• 한국컴퓨터정보학회논문지
• /
• 제21권9호
• /
• pp.19-25
• /
• 2016

This paper is to recognize and play tab chords from guitar musical sheets. The musical chord area of an input image is segmented by changing the image in saturation and applying the Grabcut algorithm. Based on a template matching, our approach detects tab starting sections on a segmented musical area. The virtual block method is introduced to search blanks over chord lines and extract tab fret segments, which doesn't cause the computation loss to remove tab lines. In the experimental tests, the prototype based classification outperforms Bayesian method and the nearest neighbor rule with the whole set of training data and its performance is similar to that of the support vector machine. The experimental result shows that the prediction rate is about 99.0% and the number of selected prototypes is below 3.0%.

• 한국BIM학회 논문집
• /
• 제12권1호
• /
• pp.72-80
• /
• 2022

It has become mandatory to designs by BIM in construction. It is urgent to make accurate decisions through the linkage between complex and various types of data in projects. In particular, block-chain based data sharing process (using BIM files, general construction submitted files) is essential to support reliable decision making in complex data flood systems. Prior to developing data sharing system based on block-chain, in this paper, a data linkage method is proposed so that practitioners can simultaneously utilize existing construction information and BIM data. Examples are shown based on the construction classification system and file expression, and incentive strategies are explored through a survey so that heterogeneous information can be used at the same time in overall projects.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2022년도 추계학술발표대회
• /
• pp.493-494
• /
• 2022

The development of livestock faces intensive farming results in a rising need for recognition of individual animals such as cows and pigs is related to high traceability. In this paper, we present a non-invasive biometrics systematic approach based on the deep-learning classification model to pig face identification. Firstly, in our systematic method, we build a ROS data collection system block to collect 10 pig face data images. Secondly, we proposed a preprocessing block in that we utilize the SSIM method to filter some images of collected images that have high similarity. Thirdly, we employ the improved image classification model of CNN (ViT), which uses the finetuning and pretraining technique to recognize the individual pig face. Finally, our proposed method achieves the accuracy about 98.66%.

• 패션비즈니스
• /
• 제9권3호
• /
• pp.64-87
• /
• 2005

The purpose at this study is to develop the appropriate bodice model for Chinese women in order to contribute to the improvement of fitness at clothing products that are exported to China. The sample group was the subject of 149 persons with the standard body shape at 19-24 years old women who reside in Beijing and Shanghai, China. The data analysis is processed for statistics using SPSSWIN 10.0 PROGRAM, and the used analysis methods are technical statistics analysis, factor analysis, group analysis, and one-way ANOVA. The outcome of this study is summarized as follows. 1. Prior to develop the tight-fitting shape of bodice model, the body classification approach by the posture and type of bending and stretching is selected to use 6 index items, and the body types are classified into bent body, right body, and pull-back body. 2. The average body size of standard body shape had 3 times of wearing experiment based on the tight-fitting shape of ESMOD bodice block drawing, and the system was corrected and supplemented to present the final bodice block drawing. 3. Comparisons have been made based on the center front line, center back line and chest circumference for each of existing bodice block for Chinese women, existing bodice block for Korean women and the combination of the bodice block under this study.