• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.1
• /
• pp.44-49
• /
• 2021

Recently, due to COVID-19, studies have been popularly worked to apply neural network to mask wearing automatic detection system. For applying neural networks, the 1-stage detection or 2-stage detection methods are used, and if data are not sufficiently collected, the pretrained neural network models are studied by applying fine-tuning techniques. In this paper, the system is consisted of 2-stage detection method that contain MTCNN model for face recognition and ResNet model for mask detection. The mask detector was experimented by applying five ResNet models to improve accuracy and fps in various environments. Training data used 17,217 images that collected using web crawler, and for inference, we used 1,913 images and two one-minute videos respectively. The experiment showed a high accuracy of 96.39% for images and 92.98% for video, and the speed of inference for video was 10.78fps.

• Korean Journal of Remote Sensing
• /
• v.38 no.1
• /
• pp.73-82
• /
• 2022

Marine fisheries resources face major anthropogenic threat from unregulated fishing activities; thus require precise detection for protection through marine surveillance. Korea developed an efficient land-based small fishing vessel monitoring system using real-time V-Pass data. However, those data directly do not provide information on fishing activities, thus further efforts are necessary to differentiate their activity status. In Korea, especially in Busan, longlining is practiced by many small fishing vessels to catch several types of fishes that need to be identified for proper monitoring. Therefore, in this study we have improved the existing fishing status classification method by applying Hidden Markov Model (HMM) on V-Pass data in order to further classify their fishing status into three groups, viz. non-fishing, longlining and other types of fishing. Data from 206 fishing vessels at Busan on 05 February, 2021 were used for this purpose. Two tiered HMM was applied that first differentiates non-fishing status from the fishing status, and finally classifies that fishing status into longlining and other types of fishing. Data from 193 and 13 ships were used as training and test datasets, respectively. Using this model 90.45% accuracy in classifying into fishing and non-fishing status and 88.23% overall accuracy in classifying all into three types of fishing statuses were achieved. Thus, this method is recommended for monitoring the activities of small fishing vessels equipped with V-Pass, especially for detecting longlining.

• Journal of Engineering Education Research
• /
• v.7 no.2
• /
• pp.5-21
• /
• 2004

The study aims to propose plans to give credit approval to those who obtain authorized civil qualifications, in accordance with the enforcement regulations under the Clause 7, Article 4 of $\ulcorner$the law on credit approval and others$\lrcorner$. Preceding studies on the grounds and principles of credit approval, analyses on the related references and materials, and surveys asking the managers of authorized civil qualifications their opinion over giving credit approval to authorized civil qualifications were conducted as the methodology of this study. Besides, a conference inviting experts from the relevant fields was held to specifically overview the contents and levels to be examined by qualification items, to conduct a face-to-face survey on directions to take in the credit approval of authorized civil qualifications, and to analyze the level and the degree of the difficulty of questions in the examinations of authorized civil qualifications. The contents and the level of credit approval in this study are as follows. For the authorized civil qualification items unable to formulate criteria in accordance with the principles of credit approval taken in the national technique qualification and other national qualifications, two factors were put under consideration for setting the level of the credit approval. First, the level and scope of work were investigated. Second, the content of qualification was compared with the course work of college. The degree of difficulty in the scope and performance of work was reviewed by specialized qualification and general qualification, respectively. Specialized qualification indicates whether or not the required knowledge and technique are acquired for performing duty in specific work fields. It falls into service fields and qualification items except qualification items on general clerical work of the national technique qualification and other national qualifications. To the contrary, general qualification is to prove the degree of acquisition of knowledge and technique for improving the basic competencies throughout diverse types of occupations. It includes competencies to verify language proficiency, mathematical and statistical capacity, problem settlement, negotiation and communication skills. When the authorized civil qualification came under the specialized qualification, the level of qualification was determined in comparison with the level of work of national qualifications. In the case of the general qualification, the credit to be approved was settled by conducting a comparative analysis on the course work of college.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.48 no.4
• /
• pp.76-85
• /
• 2011

In this paper, we propose a texture feature-based language identification using Gabor feature and wavelet-domain BDIP (block difference of inverse probabilities) and BVLC (block variance of local correlation coefficients) features. In the proposed method, Gabor and wavelet transforms are first applied to a test image. The wavelet subbands are next denoised by Donoho's soft-thresholding. The magnitude operator is then applied to the Gabor image and the BDIP and BVLC operators to the wavelet subbands. Moments for Gabor magnitude image and each subband of BDIP and BVLC are computed and fused into a feature vector. In classification, the WPCA (whitened principal component analysis) classifier, which is usually adopted in the face identification, searches the training feature vector most similar to the test feature vector. Experimental results show that the proposed method yields excellent language identification with rather low feature dimension for a document image DB.

• The Journal of Korean Institute for Practical Engineering Education
• /
• v.4 no.2
• /
• pp.98-103
• /
• 2012

Koreatech has recently adopted a long-term co-op program called IPP (Industry Professional Practice) to address problems in engineering education of Korea, but it is anticipated to face many difficulties in implementing the program due to lack of relevant experiences in Korea. In this regard, a performance evaluation scheme is required to improve the operational efficiency and judge the effectiveness of the IPP program at the same time. This study aims to develop a comprehensive performance evaluation model for the Koreatech co-op program on the basis of Kirkpatrick's four stage performance evaluation model for training programs. For this purpose, thorough review on the program evaluation literature and in-depth analyses of overseas cases of co-op performance evaluation were also conducted. The study is expected to help enhance the effectiveness of the IPP program and promote buy-in of a range of stakeholders for the first long-term co-op program in Korea.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.38A no.6
• /
• pp.486-491
• /
• 2013

This paper proposes a fast and efficient Haar-like feature selection algorithm for training classifier used in object detection. Many features selected by Haar-like feature selection algorithm and existing AdaBoost algorithm are either similar in shape or overlapping due to considering only feature's error rate. The proposed algorithm calculates similarity of features by their shape and distance between features. Fast and efficient feature selection is made possible by removing selected features and features with high similarity from feature set. FERET face database is used to compare performance of classifiers trained by previous algorithm and proposed algorithm. Experimental results show improved performance comparing classifier trained by proposed method to classifier trained by previous method. When classifier is trained to show same performance, proposed method shows 20% reduction of features used in classification.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.3
• /
• pp.333-340
• /
• 2022

The COVID-19 has made everyone anxious and people need to keep their distance. It is necessary to conduct collective assessment and screening of college students' mental health in the opening season of every year. This study uses and trains a multi-layer perceptron neural network model for deep learning to identify facial emotions. After the training, real pictures and videos were input for face detection. After detecting the positions of faces in the samples, emotions were classified, and the predicted emotional results of the samples were sent back and displayed on the pictures. The results show that the accuracy is 93.2% in the test set and 95.57% in practice. The recognition rate of Anger is 95%, Disgust is 97%, Happiness is 96%, Fear is 96%, Sadness is 97%, Surprise is 95%, Neutral is 93%, such efficient emotion recognition can provide objective data support for capturing negative. Deep learning emotion recognition system can cooperate with traditional psychological activities to provide more dimensions of psychological indicators for health.

• Journal of People, Plants, and Environment
• /
• v.23 no.3
• /
• pp.377-385
• /
• 2020

Background and objective: In the Republic of Korea, military service is mandatory. Some of new recruits have the stress from the special environment, which could cause psychological maladjustment. The military forces have operated education programs such as green camp and healing camp. The study was conducted to investigate changes in psychological and physiological stress by conducting forest healing activities along with plant scent treatment for soldiers participating in a green camp. Methods: A total of 52 soldiers were participated including maladjustment soldiers and those recommended by their military units to protect the unfit soldiers who participated in forest healing activities in the green camp. The programs that were certified by the Korea Forest Service between 2014 and 2019, and were applied for stress reduction and relaxation training were classified into stress-coping programs and forest experience programs. Post-hoc surveys and cortisol measurements were carried out. Results: Green camp soldiers experience and differences in the level of stress responses were found to be very statistically significant between the treatment and control groups. The techniques for coping with stress were not significantly different in the control group, and the treatment group showed statistically significant results. In addition, the results of analyzing changes in the concentration of cortisol and measuring physiological stress were very statistically significant in forest healing activities at 4 p.m. Conclusion: Once green camp soldiers face stressful situations, forest healing programs using forest plant scents for green camp soldiers can have positive responses and forest healing activities can reduce psychological and physiological stress responses, improving maladjustment behaviors caused by stress and positively affecting the reduction of cortisol.

• Korean Journal of Health Education and Promotion
• /
• v.15 no.1
• /
• pp.1-30
• /
• 1998

A study on skin care was conducted in order to develop proper skin care program and disseminate the Information to the consumers. The study was performed from October 2, 1997 though April 30 1997 by using questionnaires. The subjects were 1,793 from lug cities of Seoul, Pusan and, and 800 from medium and small cities of Kyongsan, Kimchon, Mokpo and KimHae cities. All subjects were females over 20 years. 1. 64.3% of the subjects said that they chose the massage packs after considering their skin condition. 55.1% of the subjects said that they did not know the side effects of the massage packs. 2. 53.3% of the subjects reported that they knew the cause of acne. and 73.3% of the respondents reported that they knew the nature of their body classified by oriental medicine. 3. The average knowledge and attitude was 10.61 :t3.46%( who it is converted to percent, it is 53.1%). The upper limit was 18.9% and lower limit was 19.0%. 4. The factors which are under the influence on knowledge of skin care were the times of massage, education level, the height of subjects, disease of skin, age, the degree of fatty body, the hour of make-up(R2=0.137). 5. The factors which are under the influence on the times of massage were education level, the experience of skin side effect, the status of physic8I health and the birth place of the subjects(R2=0.139). 6. The main factors which had significant effect on the status of face skin health were the status of physical health, economic status, age, the side effect of skin cosmetic, and the hour of make-up(R2=0.140). 7. Finally, it is recommended that training package on side effect of cosmetics, massage, physical characteristics and proper way of make-ups, and the public should be educated on the above mentioned areas.

• Journal of Environmental Health Sciences
• /
• v.35 no.6
• /
• pp.500-509
• /
• 2009

Passive samplers have been used for many years for the sampling of organic vapors in work environment atmospheres. Currently, all passive samplers used in domestic occupational monitoring are foreign products. This study was performed to evaluate variable parameters for the development of passive organic samplers, which include the geometry of the device and diffusive length for the sampler design. Four prototype diffusive lengths; A-1(4.5 mm), A-2(7.0 mm), A-3(9.5 mm), A-4(12.0 mm) were tested for adsorption performances to a chemical mixture (benzene, toluene, trichloroethylene, and n-hexane) according to the US-OSHA's evaluation protocol. A dynamic vapor exposure chamber developed and verified by related research was used for this study. The results of study are as follows. The results in terms of sampling rate and recommended sampling time test indicate that the most suitable model was A-3 (9.5 mm diffusive lengths on both sides) for passive sampler design in time weighted average (TWA) assessment. Sampling rates of this A-3 model were 45.8, 41.5, 41.4, and 40.3 ml/min for benzene, toluene, trichloroethylene, and n-hexane, respectively. The A-3 models were tested on reverse diffusion and conditions of low humidity air (35% RH) and low concentrations (0.2 times of TLV). These conditions had no affect on the diffusion capacity of samplers. In conclusion, the most suitable design parameters of passive sampler are: 1) Geometry and structure - 25 mm diameter and 490 $mm^2$ cross sectional area of diffusion face with cylindrical form of two-sided opposite diffusion direction; 2) Diffusive length - 9.5 mm in both faces; 3) Amount of adsorbent - 300 mg of coconut shell charcoal; 4) Wind screen - using nylon net filters (11 ${\mu}m$ pore size).