• 정보처리학회논문지B
• /
• 제16B권4호
• /
• pp.263-270
• /
• 2009

히스토그램 평활화는 주어진 입력 영상의 누적분포함수 CDF (Cumulative Distribution Function)를 이용하여 영상의 동적영역 (Dynamic Range)을 확장하고 히스토그램의 분포를 균등하게 함으로써 명암비를 개선한다. 그러나 히스토그램 평활화는 영상의 밝기를 과도하게 변하게 하는 단점이 있다. 본 논문에서는 과도한 명암비 향상을 억제하기위해 서브-히스토그램의 면적비 기반의 히스토그램 재분배를 이용한 적응형명암비 향상 알고리즘. 제안한 알고리즘은 영상의 동적영역을 입력영상의 휘도 평균값을 기반으로 분할하고, 분할된 영역의 면적비에 따라 밝기 분포를 재분배함으로써 과도한 밝기 변화를 효과적으로 억제 할 수 있다. 실험결과를 통하여 시각적으로 색의 왜곡이 없는 자연스러운 영상을 확인하였고, 평균값의 비교를 통해 과도한 밝기 변화를 억제한 것을 확인 할 수 있었다. 또한 히스토그램의 분포에 상관없이 대부분의 영상에서 우수한 결과를 나타내는 것을 실험결과에서 알 수 있었다.

• 무역학회지
• /
• 제44권2호
• /
• pp.269-282
• /
• 2019

This study analyzes the effects of participation in the global production network on the income inequality using panel data from 2005 to 2016 for 63 countries. In this study were used fixed effects model with autocorrelation, random effect model with autocorrelation and the GLS method. Results are as follows: First, the economic development level supports the Kuznets hypothesis. And then, the forward participation in global value chains increased income inequality, and the backward participation decreased income inequality. In order to derive more detailed estimation results, we analyzed OECD countries and non-OECD countries. First, OECD countries featured decreased, but increased beyond a certain level as a U-shaped curve, that did not support the Kuznets hypothesis. In contrast, non-OECD countries followed the Kuznets U-curve. Second, participation in the global production network showed that both OECD and non-OECD countries featured increased income inequality. In contrast, backward participation appears to mitigate income inequality both in OECD and non-OECD countries. Finally, the ratio of labor and capital is significant in mitigating income inequality in non-OECD countries in which they feature backward participation in production networks. This can be interpreted as developing economies participate in the global production network due to increased capital accumulation and increased the labor productivity.

• Asian Pacific Journal of Cancer Prevention
• /
• 제16권8호
• /
• pp.3407-3412
• /
• 2015

Background: The aim of this study was to evaluate and compare the accuracy of diffusion-weighted imaging (DWI), apparent diffusion coefficient (ADC) value, and time-intensity curve (TIC) type analysis derived from dynamic contrast-enhanced MR imaging (DCE-MRI) in differentiating benign from malignant adnexal masses. Materials and Methods: 47 patients with 56 adnexal masses (27 malignant and 29 benign) underwent DWI and DCE-MRI examinations, prior to surgery. DWI signal intensity, mean ADC value, and TIC type were determined for all the masses. Results: High signal intensity on DWI and type 3 TIC were helpful in differentiating benign from malignant adnexal masses (p<0.001). The mean ADC value was significantly lower in malignant adnexal masses (p<0.001). An ADC value< $1.20{\times}10^{-3}mm^2/s$ may be the optimal cutoff for differentiating between benign and malignant tumors. The negative predictive value for low signal intensity on DWI, and type 1 TIC were 100%. The pairwise comparison among the receiver operating characteristic (ROC) curves showed that the area under the curve (AUC) of TIC was significantly larger than the AUCs of DWI and ADC (p<0.001 for comparison of TIC and DWI, p<0.02 for comparison of TIC and ADC value). Conclusions: DWI, ADC value and TIC type derived from DCE-MRI are all sensitive and relatively specific methods for differentiating benign from malignant adnexal masses. By comparing these functional MR techniques, TIC was found to be more accurate than DWI and ADC.

• Restorative Dentistry and Endodontics
• /
• 제24권4호
• /
• pp.585-603
• /
• 1999

Esthetic availabilities used as an esthetic restorative maternals can be determined by the optical coincidence among materials, enamel and dentin. Enamel is highly translucent. That's why esthetic materials need to correspond the close translucency of enamel. But the translucent materials are affected by the background color. So it should be predicted that the color of estorative materials depend on the any thickness and the spectral reflectance of the background on which they are placed. The object of this investigation, under above hypothesis, was to determine and analyze how they affect the final color according to the thickness, translucency and background color (white, black and dentin) fill three commercial light cured composite resins(Charisma, Spectrum TPH and Z100). And correlation was analyzed to find out the possibility of the prediction when using the certain background color and thickness of materials. Followings are the result 1. The I shade of CHA showed the lowest contrast ratio($Co_7$) while the B3 shade of Z100 showed the highest contrast ratio(p<0.05). 2. The value of $L^*$ and $b^*$ on the white and dentin background is increased with decreasing thickness. And there are significant relationships between increasing thickness and each value(R>0.085). But there is a little change of $L^*$ and $b^*$ value on the black background regardless of the thickness(p>0.05). 3. For the $a^*$ value, there was little difference in values as a function of thickness and changed irregularly regardless of thickness in all background. 4. The pattern of increasing value of $L^*$ and $b^*$ with decreasing thickness was similar to the group of white and dentin background. In both dentin one showed lesser change of value. 5. The values of $L^*a^*b^*$ measured on the different background with same thickness showed the recognizable color difference(${\Delta}E^*$>2) when the thickness was below 2.6mm. 6. Contrast ratio was increased with increasing thickness with significant relationship (R>0.9). 7. Spectral reflectance of composite resins that calculated from Kubelka-Munk equation was showed little difference compared with observed value w1th decreasing thickness.

• 정보처리학회논문지B
• /
• 제11B권7호
• /
• pp.779-784
• /
• 2004

본 논문에서는 평판디스플레이 상에서 밝기 보존을 이용한 콘트라스트 향상 기법을 제안하였다. 제안한 방식은 실시간 처리를 위하여 선형추정방법을 적용하여 누적분포함수를 계산함으로써 하드웨어 복잡성을 감소시켰다. 본 논문에서는 제안한 알고리즘을 효과적으로 처리하기 위하여 누적분포함수의 샘플 값과 배럴쉬프트(Barrel Shift)를 사용하였다. 제안된 방식에 의한 처리 결과와 원 영상의 화질 평가를 위하여 시각적 검증과 히스토그램 편차를 도입하였다.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2007년도 춘계학술발표대회
• /
• pp.326-327
• /
• 2007

We investigated whether the CT images of hepatic lesions could be analyzed by computer-aided diagnosis (CAD) tool. We retrospectively reanalyzed 14 liver CT images (10 hepatocellular cancers and 4 benign liver lesions; patients who presented with hepatic masses). The hepatic lesions on CT were segmented by rectangular ROI technique and the morphologic features were extracted and quantitated using fractal texture analysis. The contrast enhancement of hepatic lesions was also quantified and added to the differential diagnosis. The best discriminating function combining the textural features and the values of contrast enhancement of the lesions was created using linear discriminant analysis. Textural feature analysis showed moderate accuracy in the differential diagnosis of hepatic lesions, but statistically insignificant. Combining textural analysis and contrast enhancement value resulted in improved diagnostic accuracy, but further studies are needed.

• 한국광학회지
• /
• 제10권5호
• /
• pp.379-385
• /
• 1999

칼라 CRT(cathode ray tube) 모니터의 화면밝기와 명암대비 레벨에 따른 화질 특성을 연구하였다. 화면밝기와 명암대비 레벨을 변화시키면서 SQRI(square root integrals)를 측정하여 시감효과를 고려한 분해능을 조사하였다. 또한 이들 레벨에서 삼색 채녈의 색도 좌표와 휘도를 측정하여 채널의 색도 좌표 일정, 색재현 영역, 그리고 채널의 독립을 만족하는 흰색의 최대 휘도를 분석하였다. 결과를 바탕으로 최적의 분해능과 색재현 특성을 가지는 화면 밝기와 명암대비 레벨을 제시하였다.

• 한국컴퓨터정보학회논문지
• /
• 제21권11호
• /
• pp.39-47
• /
• 2016

In this paper, we propose an approach for real time computation of rayleigh CLAHE using a FPGA. The contrast enhancement technique should be applied in thermal equipment having a low contrast image. And thermal equipment must be processed in real time. The CLAHE is an improved algorithm based Histogram Equalization, but the HW design is complex. A value greater than a given threshold in CLAHE should be equally distributed on the other histogram bin, this process requires iterations for the distribution. But implementation of this processing in the FPGA is constrained, so this section was implemented on the assumption of the histogram distribution or modified the operation process or implemented separately in the CPU. In this paper, we designed a distinct redistribution operation in two stages. So FPGA was designed for easy, this was designed to be distributed evenly without the assumptions and constraints. In addition, we have designed a CLAHE with the rayleigh distribution to the FPGA. The simulation shows that the proposed method provides a better image quality in the thermal image.

• 유통과학연구
• /
• 제12권12호
• /
• pp.107-118
• /
• 2014

Purpose - Due to the recent slowdown of growth in global luxury brands, which have been an engine of our domestic fashion market growth, there is an increasing need to develop a global market for domestic high-priced fashion brands. In spite of the large scale of trade between Korea and China with respect to fashion products, current trade concentrates on middle- and low-priced products rather than high-priced products. Diversification of the trade between Korea and China in terms of the price levels of trade products is needed. An understanding of Chinese consumption values and purchase intentions for luxury fashion brands will be very helpful for the establishment of strategies with the aim of increasing the level of trade with regard to high-priced fashion products. Therefore, the study aimed to identify the differences in the clothing consumption values of South Korean and Chinese women in their 20s, especially with reference to how those values affect their preference and purchase intention for luxury fashion brands. Research design, data, methodology - The study was implemented through a descriptive survey method using a self-administered questionnaire. The sample consisted of 283 Korean and 306 Chinese women in their 20s, residing in Seoul, Beijing, Guangzhou, and Shanghai. Data were collected from March 3 to 15, 2014. A total of 589 completed responses were analyzed. Data were analyzed by factor analysis, t-test, and multiple regression analysis. Results - Fivefactors for clothing consumption values were formulated: conspicuous/social value, fashionability value, personality expression value, hedonic value, and practical value. There were significant differences between Korean and Chinese consumers in clothing consumption values (except social value), preference, and purchase intention for luxury fashion brands. With regard to clothing consumption values, Chinese women attached more importance to fashionability value, personality expression value, hedonic value, and practical value than South Korean women. In addition, Chinese women's preference and purchase intention for luxury fashion brands were higher than South Korean women's. Second, in the case of South Korean women, social value, practical value, and fashionability value had positive effects on preferences for luxury fashion brands, while attractive/personality expression value and hedonic value had no influence. In contrast, in the case of Chinese women, social value and fashionability value had positive effects on preference for the luxury fashion brands. Third, in the case of South Korean women, social value and practical value had positive effects on purchase intentions for luxury fashion brands. In contrast, in the case of Chinese women, social value, personality expression value, and fashionability value had positive effects on purchase intention for luxury fashion brands. Conclusions - Considering the findings of the study, it is clear that differentiated marketing strategies are needed for luxury fashion brand markets in Korea and China. The results of the study could provide useful information that will help increase the effectiveness of luxury fashion brand marketing strategies in Korea and China.

• 핵의학기술
• /
• 제12권3호
• /
• pp.235-240
• /
• 2008

PET/CT 검사의 사용 초기에 CT는 주로 감쇠보정(Attenuation Correction: AC)의 목적으로 사용되어졌지만, CT의 성능이 향상됨으로써 조영제를 이용한 CT검사를 진단에 반영하여 보다 진보된 진단적 가치를 가질 수 있게 되었다. 조영제의 사용이 없이도 CT를 판독할 수 있지만 병변의 정확한 범위를 확인하고 정상 구조물을 구별하는데 있어서 조영제의 사용은 판독자로 하여금 판독을 매우 용이하게 하는 도움을 줄 수 있다. 하지만, 그동안 PET/CT검사 시 조영제가 감쇠보정에 영향을 줄 수 있다는 논쟁으로 이견이 많았다. 40-140 keV 정도의 낮은 엑스선을 이용하는 CT영상에 비하여, 조영제를 사용한 CT영상에서는 조영제로 인해 감쇠가 많이 되지만 511 keV의 에너지를 가진 감마선은 조영제로 인해 거의 영향을 받지 않게 되고 이로 인해 감쇠보정 시 과보정을 하게 되어 PET영상에서 오류를 나타낼 수 있다는 의견이 보고되기도 하였다. 이와는 반대로 조영제가 감쇠보정에 과대평가를 가져올 수 있다는 의견과 과보정으로 인한 표준화섭취계수에 변화의 가능성은 있으나 결정적인 영향을 미치지 않는다는 의견도 제시되었다. 본 연구에서는 조영제의 영향이 SUV에 어떠한 영향을 미치는지에 대해서만 비교 평가하였다. 2007년 12월에서 2008년 6월 사이에 본원에서 PET/CT 검사를 시행한 환자 중 요오드 조영제에 대한 부작용이 없고 당뇨병이 없는 진행성 암 환자 30명을 대상으로 하였으며, DSTe (General Electric Healthcare, Milwaukee, MI, USA)를 사용하여, 각각의 환자는 조영제를 사용하지 않은 CT를 시행한 후 PET영상을 얻었고 그 후, 조영제를 사용한 CT검사를 하였다. 각각의 CT정보로 감쇠보정을 실시, 종류별로 PET영상을 획득하여, 각 영상에 동일한 관심영역(Region of Interests : ROIs)을 설정 후, SUV를 비교 하였다. 얻어진 두 가지 결과 값의 정량 분석의 비교를 위해서 대응표본 T-검정 (Paired t-test)을 사용하였다. 검사를 시행한 30명의 환자에게서 폐, 간, 심장의 $SUV_{max}$ 값과 $SUV_{mean}$값을 측정하여 총 180개 영역을 분석하였다. 조영제를 사용하기 전과 후를 비교하였을 때, 측정한 거의 모든 영역에서 조영 후의 $SUV_{max}$와 $SUV_{mean}$가 상승하였고 통계적으로도 유의한 것으로 나타났다(p value<0.05). 심장 영역에서 조영 전 보다 조영 후의 $SUV_{mean}$값이 증가된 것으로 나타났지만, 통계적으로 유의하게 나타나지는 않았다. 조영제를 사용한 CT영상의 SUV값이 과보정되어, 조영제를 사용하지 않은 CT의 SUV 보다 높게 나왔으나 기존의 보고된 논문에서는 이와 같은 결과는 실제 임상 판독에 있어서 큰 영향을 미치지 않는다고 보고되어 있다. 그러나 원발성병변의 진행을 알아보는 과정에서는 SUV의 작은 변동도 분명 영향을 줄 수 있으므로 간 영역과 같이 SUV의 변동이 다른 영역에 비해 상대적으로 큰 영역에서는 수치의 변화에 대한 주의가 요구 될 것으로 사료된다.