• 대한방사선치료학회지
• /
• 제3권1호
• /
• pp.27-36
• /
• 1989

In order to calculate the dose on each interest point in five types of irregularly shaped fields used commonly in radiotherapy, the tissue-air ratios (TAR) in these fields for Go-60 gamma radiation were calculated using the newly devised SAR-chart. The TARs calculated from newly method of using the SAR-chart, computer method and approximation method at the interest point were compared to the TARs obtained from measurement. The result are as follows; In case of the interest points on central axis the calculated TARs in irregularly shaped fields by the above mentioned methods were well agreed within the error of $1\%$, whereas for the interest points on off-axis the calculated TARs were resulted in the maximum errors of $2.4\%,\;2.3\%$ and $8.8\%$ respectively. From these results, the accuracy of calculation method of using the SAR-chart was comfirmed.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2003년도 ICCAS
• /
• pp.118-121
• /
• 2003

In the conventional boundary estimation in EIT (Electrical Impedance Tomography), the interface between anomalies and background is expressed in usual as Fourier series and the boundary is reconstructed by obtaining the Fourier coefficients. This paper proposes a method for the boundary estimation, where the boundary of anomaly is approximated as the interpolation of front points located discretely along the boundary and is imaged by tracking the points in the region of interest. In the solution to the inverse problem to estimate the front points, the multi-layer neural network is introduced. For the verification of the proposed method, numerical experiments are conducted and the results indicate a good performance.

• 가정과삶의질연구
• /
• 제26권5호
• /
• pp.377-390
• /
• 2008

This paper investigated materialism inclination of the parents recognized by the adolescents and analysed the relationships between 5 categorized types about parents' materialism inclination and financial characteristics of the adolescents. Statistical analysis was achieved by using Cronbach'$\alpha$, paired t-test, one-way ANOVA, two-way ANOVA, Duncan's multiple range test, chi-square analysis, and Ward' hierarchical cluster analysis with a total of 801 questionnaires. The results are summarized as follows: First, the adolescents felt that calculated points of father and mother's materialism inclination were lower than median points and mother's inclination for materialism was higher than that of father. Second, five categories were classified according to materialism inclination of parents. First classified category was the type which showed both parents had little interest in materials and the second category was the type which father had strong interest in materials. The third one was the type which mother had strong interest in materials. The last two categories were the ones which both parents had much interest in materials. Final result of this study revealed that attitudes of the adolescents for materialism and money were higher for those whose parents had strong interest in materialism than those whose parents had little materialism interest. They showed not only the behaviors of impulsive and overspending consumption but also low financial satisfaction and high financial stress.

• 방송공학회논문지
• /
• 제17권5호
• /
• pp.808-816
• /
• 2012

얼굴 인식은 여러 분야에서의 활발한 연구를 통해 많은 발전이 있었고, 현재도 활발한 연구가 진행되고 있다. 최근 들어 물체 인식에 주로 사용되어온 특징점 추출 알고리즘이 얼굴 인식에도 적용되고 있다. 본 논문은 대표적인 특징점 추출 알고리즘인 SURF를 이용한다. 사람은 얼굴의 형태 및 구조가 유사하므로 물체를 인식하는 경우보다 분별력이 떨어지기 때문에 SURF를 이용한 얼굴인식의 정확도는 낮은 편이다. 이를 개선하고자 본 논문에서는 SURF를 통해 추출한 특징점에서 Gabor 웨이블릿 변환을 사용해 기술어를 추출하는 얼굴 인식 방법을 제안한다. 실험 결과에서 제안하는 방법이 기존 SURF 기반의 얼굴 인식에 비해 정확도가 약 23% 향상된 것을 확인하였다.

• Journal of applied mathematics & informatics
• /
• 제27권1_2호
• /
• pp.69-78
• /
• 2009

The accuracy of wavelet approximation at resolution h = $2^{-k}$ to a smooth function f is limited by O($h^M$), where M is the number of vanishing moments of the mother wavelet ${\psi}$; that is, the approximation order of wavelet approximation is M - 1. High accuracy points of wavelet approximation are of interest in some applications such as signal processing and numerical approximation. In this paper, we prove the scaling and translating properties of high accuracy points of wavelet approximation. To illustrate the results in this paper, we also present two examples of high accuracy points of wavelet approximation.

• 한국항행학회논문지
• /
• 제16권2호
• /
• pp.394-400
• /
• 2012

SURF(Speeded Up Robust Features)는 다양한 상태 변화에 강인한 기술자 추출 방법으로 객체 인식과 같은 분야에서 유용하게 사용되는 알고리즘이다. 이 알고리즘은 대표적인 특징점 추출 알고리즘인 SIFT(Scale Invariant Feature Transform)와 비슷한 성능을 보이면서도 수행 시간이 훨씬 빠르다는 장점이 있다. 하지만 이러한 기술자들은 회전 불변한 특징 보장을 위해서, 추출한 특징점 간의 위치 정보를 고려하지 않는다. 또한, 원본 영상을 흑백 영상으로 변환하여 사용하기 때문에, 원본 이미지의 색상 정보도 이용하지 않는다. 본 논문에서는 특징점들 간의 상대적인 위치 정보 및 색상 정보를 이용하여 SURF 기술자의 정합 성능을 개선하는 방안을 제안한다. 상대적인 위치 정보는 특징점들의 중심을 연결하는 선분과 특징점 중심에서부터 생성되는 orientation 선분 사이의 각을 기반으로 한다. 색상 정보의 경우 각 특징점이 포함하고 있는 영역에 대해 color histogram을 생성하여 사용한다. 실험을 통하여 제안된 기법의 성능 개선을 보인다.

• 전자공학회논문지CI
• /
• 제48권3호
• /
• pp.46-53
• /
• 2011

본 논문에서는 대표적인 특징점 추출 알고리즘인 SURF (Speeded Up Robust Features)를 이용한 얼굴 인식 방법을 소개한 다. 일반적으로, SURF를 이용한 물체 인식은 특징점 추출 및 정합만을 수행하지만, 본 논문에서 제안하는 SURF를 이용한 얼굴 인식 방법은 특징점 추출 및 정합뿐만 아니라 얼굴 영상 회전 및 특징점 검증을 추가로 수행한다. 얼굴 영상 회전은 특징점의 수를 증가시키기 위해 수행되며, 특징점 검증은 정확하게 정합된 특징점들을 찾기 위해 수행된다. 비록 본 논문에서 제안한 SURF를 이용한 얼굴 인식 방법은 PCA를 이용한 방법보다 연산 시간이 더 요구되었지만, 인식률은 보다 더 높았다. 이러한 실험 결과를 통해, 특징점 추출 알고리즘도 얼굴 인식에 적용할 수 있음을 확인할 수 있었다.

• 한국주거학회:학술대회논문집
• /
• 한국주거학회 2006년도 추계학술발표대회 논문집
• /
• pp.253-259
• /
• 2006

Since townscape planning zones were introduced in 2000 by a revised Urban Planning Law, interest and understanding on townscape have been enhanced in Korea. Driven by this trend of the time, interest in townscape planning has been growing in the area of urban planning, and local governments have been expressing their interest in townscape management. This study is intended to provide criteria for setting goals of constructing desirable urban environments in harmony with existing condition in Chung-ju City. To this end, questionnaire surveys and awareness research were administered and following findings were obtained in relation to user-specific preferences on view targets and view points for townscape planning in Chung-ju City. In general, citizens of Chung-ju City were found to prefer townscape frequently in use practically, while residents in other areas preferred such townscape as linked with the city' s overall reputation. Both groups were found to prefer natural landscape (Gaemyeongsan, Namsan, etc.) to artificial one as a view target, and Chung-ju Terminal and the neighborhoods and downtown(Seongseo-dong) were presumably preferred as a view point.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권9호
• /
• pp.3087-3103
• /
• 2022

Nearest neighbor query in location-based services has become a popular application. Aiming at the shortcomings of the privacy protection algorithms of traditional ciphertext nearest neighbor query having the high system overhead because of the usage of the double Hilbert curves and having the inaccurate query results in some special circumstances, a privacy protection algorithm of ciphertext nearest neighbor query which is based on the single Hilbert curve has been proposed. This algorithm uses a single Hilbert curve to transform the two-dimensional coordinates of the points of interest into Hilbert values, and then encrypts them by the order preserving encryption scheme to obtain the one-dimensional ciphertext data which can be compared in numerical size. Then stores the points of interest as elements composed of index value and the ciphertext of the other information about the points of interest on the server-side database. When the user needs to use the nearest neighbor query, firstly calls the approximate nearest neighbor query algorithm proposed in this paper to query on the server-side database, and then obtains the approximate nearest neighbor query results. After that, the accurate nearest neighbor query result can be obtained by calling the precision processing algorithm proposed in this paper. The experimental results show that this privacy protection algorithm of ciphertext nearest neighbor query which is based on the single Hilbert curve is not only feasible, but also optimizes the system overhead and the accuracy of ciphertext nearest neighbor query result.

• 치위생과학회지
• /
• 제11권1호
• /
• pp.1-6
• /
• 2011

본 연구는 인문계 고등학생의 주관적 구강건강관심도와 인지도에 따른 구강건강 실천도를 파악하여, 올바른 구강 건강 습관형성을 유도하고 실천하기 위한 정보를 제공하고자 2010년 5월 20일부터 6월 20일까지 전라북도에 소재하고 있는 일부 인문계 고등학교 남학생을 대상으로 편의표본추출법에 의한 자기기입식의 설문조사를 실시하였다. 수집된 자료는 SPSS 12.0 프로그램을 이용하여 통계분석을 실시한 결과는 다음과 같다. 1. 주관적 구강건강관심도는 관심이 있다가 52.3%, 보통이다는 32.6%, 관심없다는 14.9%였고, 인지도는 건강하다가 42.6%, 보통이다가 44.9%, 건강하지 않다는 12.7%였다. 2. 치과방문경험은 92.6%이었으며, 방문이유는 충치치료가 69.2%로 가장 많았다. 3. 구강건강 관심도에 따른 구강건강실천 요인 중 잇솔질이 3.89점으로 가장 높았고, 교육 및 관심은 3.18점, 식이조절은 2.93점, 정기적인 방문은 2.69점, 구강위생용품의 실천은 2.12점의 순이었다. 4. 주관적 구강건강인지도에 따른 구강건강실천 요인 중 잇솔질이 3.89점으로 가장 높았고, 교육 및 관심은 3.17점, 식이조절은 2.93점, 정기적인 방문은 2.69점, 구강위생용품의 실천은 2.12점의 순이었다. 5. 주관적 구강건강인지도와 구강건강실천 요인과의 상관관계에서 주관적 구강건강인지도를 건강하다고 인지할수록 구강건강 실천도가 높았다(p<0.01). 6. 주관적 구강건강인지도에 구강건강실천 요인이 미치는 영향은 잇솔질을 실천(p<0.01)과 정기적으로 치과를 방문(p<0.05)하는 집단에서 주관적 구강건강인지도가 높았다.