• Journal of the Society of Disaster Information
• /
• v.20 no.2
• /
• pp.302-314
• /
• 2024

Purpose: This study proposes a fall detection model based on a top-down deep learning pose estimation model to automatically determine falls of multiple workers in an underground utility tunnel, and evaluates the performance of the proposed model. Method: A model is presented that combines fall discrimination rules with the results inferred from YOLOv8-pose, one of the top-down pose estimation models, and metrics of the model are evaluated for images of standing and falling two or fewer workers in the tunnel. The same process is also conducted for a bottom-up type of pose estimation model (OpenPose). In addition, due to dependency of the falling interference of the models on worker detection by YOLOv8-pose and OpenPose, metrics of the models for fall was not only investigated, but also for person. Result: For worker detection, both YOLOv8-pose and OpenPose models have F1-score of 0.88 and 0.71, respectively. However, for fall detection, the metrics were deteriorated to 0.71 and 0.23. The results of the OpenPose based model were due to partially detected worker body, and detected workers but fail to part them correctly. Conclusion: Use of top-down type of pose estimation models would be more effective way to detect fall of workers in the underground utility tunnel, with respect to joint recognition and partition between workers.

• Journal of Food Hygiene and Safety
• /
• v.25 no.2
• /
• pp.162-169
• /
• 2010

The compulsory beef labelling system has launched from January 1st 2007 by the amended Food Hygiene Law, we were checked the actual conditions of beef origin with a nationwide scale by the Hanwoo differentiation specific test method which was developed by Korea FDA using 90 SNP biomarkers. The test method is useful tool to differentiate the beef origin carrying out the mission of KFDA's annual food safety management guidance. Also we have technically transferred the Hanwoo differentiation specific test method to other institutes as well regional KFDA and established the training program as a regular course in Korea Human Resource Development Institute for Health and Welfare. The beef used in this study were collected according to the 2009 Food safety guidance in roast beef restaurants where business site area greater than 100 $m^2$. Total 216 samples were consisted of 48 samples of the Seoul area and 168 of the region. The monitoring result from restaurants in all the region of Korea showed that 3 of 216 Hanwoo-labelled beefs were found out as a non-Hanwoo (1.3%). This results are gradually deceasing trend compared with 34.0% in 2005, 30.1% in 2006, 3.2% in 2007 and 5.14% in 2008. From these data, the Hanwoo differentiation specific test method on the settlement of the compulsory beef labelling system has an important role. As a outcome of this project, we might be considered the early settlement of the compulsory beef labelling system, technically transferred to other institutes and the establishment of regular training program of the test method.

• Journal of The Korean Association For Science Education
• /
• v.40 no.3
• /
• pp.347-358
• /
• 2020

He purpose of this study is to design and apply a pilot assessment tools for scientifically gifted and talented elementary school students with lower grades. The assessment tool consists of three parts: homeroom teachers' recommendation, paper and pencil test and performance tests. The tools are verified whether they are suitable for unique characteristics of young children and enable to attract active participation. For suitability check, students' performance tests were inductively analyzed and 30 behavioral patterns were shown which were included and partially exceeded the level of lower elementary students' performance expectation in NGSS. As a result, we concluded that assessment tool developed in this study will be effective in discriminating young pupils' scientific talents. Then for participation check, we compared the number of coding references as an indicator of participation. Two cases were found that students with high interest participated passively in performance tests. We found these 'passive participants' had excessive scientific experiences and extremely narrow region of interest, during the process of complex interpretation between the results of this assessment tool and in-depth interviews with homeroom teachers. We found out in this study that newly developed tools can be used in school scene after modifying and elaboration through accumulation of more case studies.

• Food Science of Animal Resources
• /
• v.24 no.2
• /
• pp.151-155
• /
• 2004

This study was performed to develop the method of differentiation fresh and frozen meat by using the measurement of mitochondrial malate dehydrogenase. The principle of this experiment is based on the fact the enzyme proteins associated with mitochondria membrane could be released by freezing. The methods were studied by measurements of protein concentration of meat press juice, WHC (water-holding capacity), drip loss and mitochondrial malate dehydrogenase enzyme activity. Samples were stored at 4$^{\circ}C$ and -18$^{\circ}C$ during storage period, respectively. Protein concentration of meat press juice was ranged from 8.5 mg/mL to 12.7 mg/mL and increased by freezing below at -18$^{\circ}C$(p<0.05). The WHC was not significantly different between fresh meat and frozen chicken meat (p＞0.05). The amount of drip loss of fresh and frozen chicken meat at 4$^{\circ}C$ and -18$^{\circ}C$ was not significantly different (p＞0.05). Mitochondrial malate dehydrogenase activity of frozen meat (-18$^{\circ}C$) was significantly higher (p＜0.05) than that of fresh meat. Also, enzyme activity of frozen meat was maintained at the same level after 3 minutes reaction. But fresh meat had not this reaction. From these results, it suggests that mitochondrial malate dehydrogenase can be used as a promising enzyme to differentiate between fresh and frozen meat.

• Journal of Korean Society of Forest Science
• /
• v.96 no.4
• /
• pp.387-392
• /
• 2007

Characteristics of needle morphology and anatomy were examined in 14 populations of Abies nephrolepis (Trautv.) Maxim. and A. koreana Wilson. Additionally we studied the classification index to distinguish between the species by the method of discriminant analysis. Characteristics of needle for A. nephrolepis could be distinguished from those for A. koreana by flatten arrangement, thin and long length for needle form, many stomata row, and marginal position of resin duct Nested ANOVA showed that there were statistically significant differences among populations as well as among individuals within populations in all 9 needle traits. For the needle indices such as needle thickness, number of stomata row, and the distance between resin duct and vascular for both species, variance components among populations were larger than those among individuals within populations. The characteristics that contributed most to the separation of A. nephrolepis and A. koreana according to the discriminant analysis using stepdisc procedures were needle index and thickness of needle, needle arrangement index, distance between resin duct and vascular, and number of stomata row.

• Phonetics and Speech Sciences
• /
• v.15 no.1
• /
• pp.25-34
• /
• 2023

This study examined the utility of the acoustic features of vowels as cues for the place of articulation of Korean nasal consonants. In the acoustic analysis, spectral and temporal parameters were measured at the 25%, 50%, and 75% time points in the vowels neighboring nasal consonants in samples extracted from a spontaneous Korean speech corpus. Using these measurements, linear discriminant analyses were performed and classification accuracies for the nasal place of articulation were estimated. The analyses were applied separately for vowels following and preceding a nasal consonant to compare the effects of progressive and regressive coarticulation in terms of place of articulation. The classification accuracies ranged between approximately 50% and 60%, implying that acoustic measurements of vowel intervals alone are not sufficient to predict or classify the place of articulation of adjacent nasal consonants. However, given that these results were obtained for measurements at the temporal midpoint of vowels, where they are expected to be the least influenced by coarticulation, the present results also suggest the potential of utilizing acoustic measurements of vowels to improve the recognition accuracy of nasal place. Moreover, the classification accuracy for nasal place was higher for vowels preceding the nasal sounds, suggesting the possibility of higher anticipatory coarticulation reflecting the nasal place.

• Korea journal of population studies
• /
• v.29 no.3
• /
• pp.89-109
• /
• 2006

The 'Joongsancheung(JSC)', a unique term for the middle class in Korea, is defined as a stratum sharing common lifestyles and a certain level of life chances. It involves non-economic factors such as life chance, educational attainment, occupational groups as well as economic factor. Such objective measures as the occupational status of the main breadwinner, family income, and the educational level of respondent, and subjective measures of class identification are used for the operational definition of the JSC. Data from a national survey of 1,515 respondents is analyzed to investigate the change of the JSC in size and the major determinants of class identification. The results show that while there is no strong evidence of any significant change of the JSC by the objective measures during the recent decade, there seems to be a slight decrease in the subjective class identification. In addition, binary logistical regression analysis reveals that self-identification of JSC is heavily influenced by house ownership, along with subjective evaluation of one's own income and property ownership. This study demonstrates that the apparent class polarization in Korean society reflects not so much objective conditions but subjective perception of respondent of his or her circumstance. It is suggested that problems of housing and relative derivation people have as regards income and property should be resolved to alleviate such class polarization in Korean society.

• Korean Journal of Remote Sensing
• /
• v.34 no.6_1
• /
• pp.905-923
• /
• 2018

In order to generate high-resolution products using UAV images, it is necessary to analyze the sharpness of the themselves measured through image analysis. When images that have unclear sharpness of UAV are used in the production, they can have a great influence on operations such as acquisition and mapping of accurate three-dimensional information using UAV. GRD (Ground Resolved Distance) has been used as an indicator of image clarity. GRD is defined as the minimum distance between two identifiable objects in an image and is used as a concept against the GSD (Ground Sampling Distance), which is a spatial sample interval. In this study, GRD is extracted by analyzing the edge target without visual analysis. In particular, GRD to GSD ratio (GRD/GSD), or GRD expressed in pixels, is used as an index for evaluation the relative image sharpness. In this paper, GRD is calculated by analyzing edge targets at various altitudes in various shooting environments using a rotary wing. Using GRD/GSD, it was possible to identify images whose sharpness was significantly lowered, and the appropriateness of the image as an image clarity index was confirmed.

• Journal of Bio-Environment Control
• /
• v.31 no.4
• /
• pp.270-278
• /
• 2022

This study was conducted using deep learning technology to classify for 'Mihwang' peach maturity with RGB images and fruit quality attributes during fruit development and maturation periods. The 730 images of peach were used in the training data set and validation data set at a ratio of 8:2. The remains of 170 images were used to test the deep learning models. In this study, among the fruit quality attributes, firmness, Hue value, and a^* value were adapted to the index with maturity classification, such as immature, mature, and over mature fruit. This study used the CNN (Convolutional Neural Networks) models for image classification; VGG16 and InceptionV3 of GoogLeNet. The performance results show 87.1% and 83.6% with Hue left value in VGG16 and InceptionV3, respectively. In contrast, the performance results show 72.2% and 76.9% with firmness in VGG16 and InceptionV3, respectively. The loss rate shows 54.3% and 62.1% with firmness in VGG16 and InceptionV3, respectively. It considers increasing for adapting a field utilization with firmness index in peach.

• Journal of the Korea Society of Computer and Information
• /
• v.20 no.3
• /
• pp.47-56
• /
• 2015

Automated detection of interesting web pages could be used in many different application domains. Determining a user's interesting web pages can be performed implicitly by observing the user's behavior. The task of distinguishing interesting web pages belongs to a classification problem, and we choose white box learning methods (fixed effect logit regression and support vector machine) to test empirically. The result indicated that (1) fixed effect logit regression, fixed effect SVMs with both polynomial and radial basis kernels showed higher performance than the linear kernel model, (2) a personalization is a critical issue for improving the performance of a model, (3) when asking a user explicit grading of web pages, the scale could be as simple as yes/no answer, (4) every second the duration in a web page increases, the ratio of the probability to be interesting increased 1.004 times, but the number of scrollbar clicks (p=0.56) and the number of mouse clicks (p=0.36) did not have statistically significant relations with the interest.