• 정보과학회 논문지
• /
• 제44권1호
• /
• pp.44-50
• /
• 2017

보행자 검출은 수년간 광범위하게 연구된 문제이며, 자율주행 자동차와 운전자 보조시스템에서 매우 중요한 역할을 차지하고 있다. 특히, 계층적 분류기[1]와 Histogram of Gradient[2]특징벡터 등 영상기반의 보행자 검출기법과 ConvNet같이 deep model을 이용하여 검출하는 기법들이 연구되었고 검출성능은 꾸준히 상승하였다. 하지만 보행자 검출은 작은 오차에도 생명과 연관된 문제를 야기할 수 있기 때문에, 자율주행 시스템의 보행자검출 오차율은 더욱 낮출 필요가 있다. 따라서 본 연구에서는 Faster R-CNN 응용 기법에 새로 개발한 데이터 학습 모델을 적용하여 보행자 검출 오류를 줄이는 기법을 제안한다. 그리고 기존에 제안된 모델들과 비교를 통해, 보행자 검출에 있어 제안된 방법의 우수성을 보이고자 한다.

• 대한방사선기술학회지:방사선기술과학
• /
• 제41권2호
• /
• pp.109-113
• /
• 2018

Despite the continuous development of software, it is continuously pursued to enlarge the examination area of FOV in order to reduce the factors of inconsistency in images that appear in continuous examination during wide area examination using contrast agent such as whole body angiography. In this study, we investigated the optimal FOV by comparing the SNR values according to the changes of FOV. The change of the FOV was gradually changed to $270{\times}200$, $300{\times}223$, $330{\times}244$, $360{\times}266$ and $380{\times}281$. SE images at TR 450 msec and TE 10 msec, FSE images at TR 2,000 msec, TE 80 msec, and GE images were scanned at TR 117 msec, and TE 16 msec. SNR values were calculated from the mean values of signal intensities of five phantom images and the signal intensity values of four background standard deviations. As a result of the study, the signal intensity and the SNR value according to the change of the FOV value gradually increased as the FOV was increased, but it was found that the SNR value decreased at a constant size. In conclusion, the results are different from previous studies that the SNR increases as the FOV increases. The cause of these results could not be confirmed. However methods that can be imaged and included within the effective FOV should be considered.

• Journal of Acupuncture Research
• /
• 제20권4호
• /
• pp.66-84
• /
• 2003

Objectve : Recent stuides suggested that there is a strong correlation between acupuncture stimulation and its cortical activation. Another study showed that either positive or negative BOLD effects could be observed depending on anatomical structure in acupuncture. Methods : 1) Subjects and paradigms: Two separate stimulation paradigms were performed in this study. To see the effects of electro-acupuncture stimulation on K7(Fuliu), the experiment was carried out on six healthy volunteers, using the gradient echo sequence with the 3.0T whole-body MRI system(ISOL). After the needle insertion on right K7(Fuliu), 2 Hz of electric stimulation was given for 30 seconds, repeated five times, with 30 seconds' intervals. During the intervals while there was no electro-stimulation, the subjects performed motor task as a reference. The image analysis including motion correction, talairach transformation, and smoothing was done using SPM99. Results: The electro-acupuncture stimulation on K7(Fuliu) activated Brodmann's Areas of 9, 19, 23, 24, 31, 32, 39 which may be the central pathways of the electro-acupuncture stimulation on K7(Fuliu) and those brain areas may be related with the function of the electro-acupuncture stimulation on K7(Fuliu).

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제11권4호
• /
• pp.2124-2148
• /
• 2017

Accurate facial expression recognition (FER) requires reliable signal filtering and the effective feature extraction. Considering these requirements, this paper presents a novel approach for FER which is robust to noise. The main contributions of this work are: First, to preserve texture details in facial expression images and remove image noise, we improved the anisotropic diffusion filter by adjusting the diffusion coefficient according to two factors, namely, the gray value difference between the object and the background and the gradient magnitude of object. The improved filter can effectively distinguish facial muscle deformation and facial noise in face images. Second, to further improve robustness, we propose a new feature descriptor based on a combination of the Histogram of Oriented Gradients with the Canny operator (Canny-HOG) which can represent the precise deformation of eyes, eyebrows and lips for FER. Third, Canny-HOG's block and cell sizes are adjusted to reduce feature dimensionality and make the classifier less prone to overfitting. Our method was tested on images from the JAFFE and CK databases. Experimental results in L-O-Sam-O and L-O-Sub-O modes demonstrated the effectiveness of the proposed method. Meanwhile, the recognition rate of this method is not significantly affected in the presence of Gaussian noise and salt-and-pepper noise conditions.

• ALGAE
• /
• 제30권2호
• /
• pp.139-146
• /
• 2015

Urchin barrens have been a major issue of rocky coastal ecosystems in temperate regions. In South Korea, the east coast and Jeju Island have especially been a focus because the area of barren ground increases in spite of continual efforts to install artificial reefs. This study approached the urchin barrens issue in South Korea, by focusing on a correlational analysis of urchin and macroalgal abundance. Urchin density and algal species coverage were obtained using a quadrat image analysis. Subtidal sites were then classified into three groups according to the average densities of urchins to evaluate the characterization of the macroalgal community: no urchin (NU) zone; transition (TR) zone, $4inds.\;m^{-2};$ and urchin (UR) zone, ${\geq}8inds.\;m^{-2}$. The average urchin density in the study site was $4.7inds.\;m^{-2}$ and 57 macroalgal species were found in the study site. From the NU zone to UR zone, total species number, species diversity index and evenness gradually decreased, whereas the dominance index increased. The algae species with negative correlations were Grateloupia divaricata, Polysiphonia morrowii, Chondracanthus intermedius, Delesseria violacea, Desmarestia viridis; and those with positive correlations were the crustose corallines, Sargassum horneri. Other species were not significantly correlated with urchin density. The significant correlations indicate that the abundance of some macroalgal species is proportionally regulated by sea urchin density. This study also shows how macroalgal vegetation changes in response to an urchin's density gradient in a natural condition; and there is a TR zone that existed with respect to an intermediate level of algal abundance.

• 대한한방내과학회지
• /
• 제25권3호
• /
• pp.559-568
• /
• 2004

Diagnosis of Raynaud's phenomenon is primarily based on clinical symptoms. Cold stress test(CST) done by DITI(Digital Infrared Thermographic Image) can be helpful for objective diagnosis. The cold stress test was performed three times by DITI; the first after 15 minutes of rest, the second right after one minute of soaking in $20^{\circ}C$ water, the third ten minutes after immersion. For a clear diagnosis, the temperature of the finger tips must be low, or the thermal difference between the metacarpophalangeal joints and the fingertips must be large. Also the evaluation of treatment depends on decrease of thermal gradient between the metacarpophalangeal joints and the finger tips after CST. In oriental medicine Raynaud's phenomenon can be categorized by coldness of the limbs or numbness. Numbness was diagnosed as depression of Ki and Chiljehyangbuhwan(Qizhixiangfuwan) was prescribed. Positive results were observed, not only in follow up CST, but also Visual Analogue Scale after treatment.

• 디지털콘텐츠학회 논문지
• /
• 제10권4호
• /
• pp.579-585
• /
• 2009

Tyrtshnikov[9]의 연구에서는 토플리츠 선형시스템에서 토플리츠 선행조건으로 일반해를 구하는 방법들을 제시하고 있다. 또한 대칭 토플리츠 행렬에서의 선행조건 행렬을 선택하는 방법도 소개 하였다. 본 연구는 토플리츠 시스템에서 새롭게 선행조건 찾는 방법을 소개하고 있으며, 선행조건행렬들의 분석을 통해 대칭 토플리츠 행렬의 고유값들과 대칭 토플리츠행렬로 부터 생성된 선행조건행렬의 고유값들이 매우 근접하다는 결과를 나타내고 있다. 즉, 선행조건시스템 $C_0^{-1}T$의 고유값들은 1에 모두 접근하게되면, 선행조건 시스템의 수렴속도는 superlinear이다. 본 연구에서 생성된 선행조건행렬 $C_0$은 선행조건시스템의 superlinear의 수렴속도로 계산하게 된다. 또한 토플리츠 행렬은 이미지 프로세싱이나 시그널 프로세싱에서 많이 응용되고 있으므로 본 연구에서 개발한 선행조건행렬로부터 다양한 응용성을 높일 수 있다. 본연구의 또 다른 특징은 토플리츠 행렬의 중요한 성질을 보존하면서 선행조건행렬을 생성하였다.

• 한국컴퓨터정보학회논문지
• /
• 제8권4호
• /
• pp.13-20
• /
• 2003

영상 시퀀스로부터 추정된 움직임벡터분산에 관한 연구에 기초하여 본 논문에서는 고속블록정합움직임 추정을 위한 구속조건을 적용한 다이아몬드탐색(DS) 알고리즘을 제안한다. 영상 시퀀스에서의 움직임벡터가 수직 또는 수평방향으로 2개화소이내의 거리에서 탐색되어지는 점을 고려하여, DS 알고리즘은 수직 또는 수평방향의 움직임벡터의 추정을 강조함으로써 새로운 3단계 탐색알고리즘에 비하여 오차율 면에서 유사한 결과를 나타내지만 계산량을 줄일 수 있었음을 확인하였다. 또한, DFD를 구속조건으로 DS 알고리즘에 적용함으로써 움직임이 없는 안정된 블록에서의 움직임벡터의 계산에 소요되는 계산 량을 줄일 수 있을 뿐 아니라 국부 해를 예측할 가능성을 감소시킬 수 있었다. 따라서 제안된 구속조건을 적용한 DS 알고리즘은 기존은 DS 알고리즘뿐 만 아니라 4단계 탐색, 블록기반 기울기-감소 탐색에 비해 평균자승오차 또는 필요한 탐색 점의 수에서 우수한 결과를 나타냄을 알 수 있었다.

• 비파괴검사학회지
• /
• 제33권2호
• /
• pp.187-192
• /
• 2013

크랙 계면에서의 접촉 열저항에 의해 만들어지는 온도 강하 특성을 적외선을 이용하여 측정함으로써 피로균열을 검사하였다. 크랙과 같은 불연속면을 지나는 열 유동은 연속체에서와는 달리 접촉경계면에서 급격한 온도구배를 나타내는데 이 변화를 표면에서의 적외선 방사량으로 측정하여 피로균열의 위치를 탐지하였다. 표준 피로균열 시편의 중앙부에 노치를 생성한 후 저주기 피로균열을 노치 끝단으로부터 발생시킨 다음, 이 시편의 한쪽 끝단에 할로겐램프를 이용하여 일정한 열을 가함으로서 시편의 길이 방향으로 열 유동이 일어나도록 하면서 시편의 표면온도 분포를 적외선 실험장치로 계측하였다. 열 유동이 크랙을 지나면서 온도구배가 크게 변화하는 것을 적외선 열화상 이미지로부터 알 수 있었으며 균열 면에서의 적외선 온도 데이터로부터 크랙의 위치와 크기를 평가할 수 있음을 실험적으로 확인하였다.

• 천문학회보
• /
• 제43권1호
• /
• pp.42.4-43
• /
• 2018

Most galaxies are believed to evolve through mergers and accretions. In particular, minor mergers and gas accretion appear to play an important role in galaxy evolution in the present-day Universe. Tidally-disrupted debris from such processes remain as diffuse, low-surface brightness structures because the dynamical timescale in the outskirts is significantly longer than that in the central regions. Although these structures will give us useful insight into the mass assembly history of galaxies, it is difficult to detect them due to their faint surface brightness. In order to investigate the structural properties of outskirts in nearby galaxies, we conduct deep and wide-field imaging survey with KMTNet. We present our observing strategy and an optimal data reduction process to recover faint extended features in the images of KMTNet. Using the imaging data of NGC 1291 obtained from KMTNet, we find that a peak-to-peak sky gradient can be reduced less than 0.4-0.6% of the original sky level in the entire image. We also find that we can reach the surface brightness of ${\mu}_{(B,1{\sigma})}$ ~ 29.5, ${\mu}_{(R,1{\sigma})}$ ~ 28.5 mag $arcsec^{-2}$ in one-dimensional profile, that is mainly limited by the uncertainty in the sky determination. It indicates that deep imaging data of KMTNet is suitable to study the extended faint features of nearby galaxies, such as stellar halos, outer disks, and dwarf companions.