• Journal of the Institute of Electronics and Information Engineers
• /
• v.52 no.4
• /
• pp.164-174
• /
• 2015

In a real scenario, radiometric change has frequently occurred in the stereo image acquisition process using multiple cameras with geometric characteristics or moving a single camera because it has different camera parameters and illumination change. Conventional stereo matching algorithms have a difficulty in finding correct corresponding points because it is assumed that corresponding pixels have similar color values. In this paper, we present a new method based on the local descriptor reflecting intensity, gradient and texture information. Furthermore, an adaptive weight for local descriptor based on the entropy is applied to estimate correct corresponding points under radiometric variation. The proposed method is tested on Middlebury datasets with radiometric changes, and compared with state-of-the-art algorithms. Experimental result shows that the proposed scheme outperforms other comparison algorithms around 5% less matching error on average.

• Journal of Korea Multimedia Society
• /
• v.25 no.10
• /
• pp.1375-1385
• /
• 2022

In this paper, we propose a semi-supervised domain adaptation solution to deal with practical face recognition (FR) scenarios where a single face image for each target identity (to be recognized) is only available in the training phase. Main goal of the proposed method is to reduce the discrepancy between the target and the source domain face images, which ultimately improves FR performances. The proposed method is based on the Domain Adatation network (DAN) using an MMD loss function to reduce the discrepancy between domains. In order to train more effectively, we develop a novel loss function learning strategy in which MMD loss and cross-entropy loss functions are adopted by using different weights according to the progress of each epoch during the learning. The proposed weight adoptation focuses on the training of the source domain in the initial learning phase to learn facial feature information such as eyes, nose, and mouth. After the initial learning is completed, the resulting feature information is used to training a deep network using the target domain images. To evaluate the effectiveness of the proposed method, FR performances were evaluated with pretrained model trained only with CASIA-webface (source images) and fine-tuned model trained only with FERET's gallery (target images) under the same FR scenarios. The experimental results showed that the proposed semi-supervised domain adaptation can be improved by 24.78% compared to the pre-trained model and 28.42% compared to the fine-tuned model. In addition, the proposed method outperformed other state-of-the-arts domain adaptation approaches by 9.41%.

• Journal of the Korean Society of Environmental Restoration Technology
• /
• v.26 no.6
• /
• pp.117-133
• /
• 2023

This study assessed the ecological health of border areas using the ecological health index and suggested Indicators for maintaining and promoting ecological values. We analyzed the change trends, pressures, and resilience of ecosystems and services in border areas, and identified their current status and sustainability. The main findings were: (1) ecological assets and ecosystem services in border areas could be compared through ecosystem health assessment; (2) it can be used to set priorities for management and conservation by identifying the relative importance and vulnerability of ecosystems and ecosystem services in each border area; and (3) the index presented in this study can be judged to have explanatory power for the characteristics of border areas and ecosystem health when compared to previous studies. Limitations of this study include the lack of literature and statistical data at the local government level and the resulting limited application of evaluation methods, which limited direct regional comparisons. To overcome the research limitations, further studies are needed, such as establishing ecological information in border areas, mapping and assessment of ecosystem services, and developing and applying assessments that reflect the opinions and participation of various stakeholders. This study was the first attempt to assess the health of ecosystems and ecosystem services in border areas and provided an important baseline for future changes in border areas. In the future, it will be helpful in national and local government policies and ecological assetecosystem management by supplementing insufficient information and presenting clear goals.

• Journal of Nuclear Fuel Cycle and Waste Technology(JNFCWT)
• /
• v.13 no.3
• /
• pp.171-180
• /
• 2015

This study was performed on the removal of Cs, one of the main high- radioactive nuclides contained in the high-radioactive seawater waste (HSW), by adsorption with IE911 (crystalline silicotitanate type). For the effective removal of Cs and the minimization of secondary solid waste generation, adsorption of Cs by IE911 (hereafter denoted as IE911-Cs) was effective to carry out in the m/V (ratio of absorbent weight to solution volume) ratio of 2.5 g/L, and the adsorption time of 1 hour. In these conditions, Cs and Sr were adsorbed about 99% and less than 5%, respectively. IE911-Cs could be also expressed as a Langmuir isotherm and a pseudo-second order rate equation. The adsorption rate constants (k₂) were decreased with increasing initial Cs concentrations and particle sizes, and increased with increasing ratios of m/V, solution temperatures and agitation speeds. The activation energy of IE911-Cs was about 79.9 kJ/mol. It was suggested that IE911-Cs was dominated by a chemical adsorption having a strong bonding form. From the negative values of Gibbs free energy and enthalpy, it was indicated that the reaction of IE911-Cs was a forward, exothermic and relatively active at lower temperatures. Additionally, the negative entropy values were seen that the adsorbed Cs was evenly distributed on the IE911.

• Journal of Nutrition and Health
• /
• v.25 no.2
• /
• pp.179-186
• /
• 1992

Dietary intakes of 57 college women were measured by 7-day weight food records(7WR) Nutrient intakes of 7WR were compared with results of using 1-day weighted food records(1WR) and 2-day weighed food records(2WR), 7WR were recorded for 7 consecutive days by subjects and its first two days were chosen to provide 1WR and 2WR Mean nutrient intakes of the group were not significantly different between 1WR or 2WRand 7WR. Within-person variations were found to be greater than between-person variations. The ratio of within- and between-person variability was lowest for the vitamin A and highest for the calcium. Data on 7WR were analyzed to estimate the number of recording days needed to ensure r$\geq$0.9 between observed and turn mean intakes. The 11 to 13 days needed for energy sugar protein and vitamin A and 20 days for calcium The results indicated that the number of days needed varied substantially among individuals for the same nutrient and within individuals for different nutrients.