• Journal of Chest Surgery
• /
• v.31 no.2
• /
• pp.142-148
• /
• 1998

Video-assisted thoracic surgery(VATS) is emerging as a viable alternative to thoracotomy when surgical treatment of spontaneous pneumothorax is required. 20 patients with spontaneous pneumothorax underwent bullectomy between July 1995 and May 1996. The patients were divided into two groups : Control group ; the patients who received with mid-axillary approach(n=10), Experimental group ; the patients who received with VATS (n=10). The results were as follows ; 1. The total sex distribution was male predominance (male:female=17:3). Mean age of control group was 29.6$\pm$9.8 years and experimental group was 27.2$\pm$11.9 years. 2. The mean period of postoperative chest tube indwelling duration and hospital stay were 3.3$\pm$0.8 days and 7.9$\pm$1.2 days in control group and 2.1$\pm$0.9 days and 5.2$\pm$3.1 days in experimental group(p=0.005 and p=0.02). 3. The mean time of operation, vital signs and arterial blood gas analysis did not showed any statistical differences between the groups. 4. Percent recovery of tidal volume and forced vital capacity were significantly improved in experimental group comparing with control group (p<0.05). 5. The patients undergoing VATS experienced significantly less postoperative pain and limitation of motion. In conclusion, VATS is safe and offers the potential benefits of shorter postoperative hospital stays and less pain with cosmetic benefits.

• Journal of KIISE
• /
• v.42 no.4
• /
• pp.451-458
• /
• 2015

Previous multimodal learning methods focus on problem-solving aspects, such as image and video search and tagging, rather than on knowledge acquisition via content modeling. In this paper, we propose the Multimodal Concept Hierarchy (MuCH), which is a content modeling method that uses a cartoon video dataset and a character-based subtitle generation method from the learned model. The MuCH model has a multimodal hypernetwork layer, in which the patterns of the words and image patches are represented, and a concept layer, in which each concept variable is represented by a probability distribution of the words and the image patches. The model can learn the characteristics of the characters as concepts from the video subtitles and scene images by using a Bayesian learning method and can also generate character-based subtitles from the learned model if text queries are provided. As an experiment, the MuCH model learned concepts from 'Pororo' cartoon videos with a total of 268 minutes in length and generated character-based subtitles. Finally, we compare the results with those of other multimodal learning models. The Experimental results indicate that given the same text query, our model generates more accurate and more character-specific subtitles than other models.

• Journal of Chest Surgery
• /
• v.31 no.1
• /
• pp.40-45
• /
• 1998

Recently, video-assisted thoracoscopic surgery for mediastinal lesions has been considered a new effective therapeutic method. From March, 1992 to April, 1997, 33 cases of video assisted thoracoscopic surgery for mediastinal lesions were performed. Gender distribution was 16 males and 17 females. Average age was 42 years old(ranged from 14 to 69). The locations of lesions were anterior mediastinum in 14 cases, middle mediastinum in 5 cases, posterior mediastinum in 11 cases, and superior mediastinum in 3 cases. These included 9 neurilemmomas, 5 benign cystic teratoma, 4 pericardial cysts, 2 ganglioneuroma, 2 thymus, 2 thymic cyst, 1 thymoma, 2 esophageal leiomyomas, 1 dermoid cyst, 1 lipoma, 1 malignant lymphoma, 1 bronchogenic cyst, 1 pericardial effusion, and 1 Boerhaave's disease with empyema. Working window was needed in 6 cases. We converted to open thoracotomy in 6 cases. Reasons of convertion to open thoracotomy were large sized mass(1), severe adhesion(3), and difficult location to approach(2). The average operation time was 116min($\pm$56 min). The average chest tube drainage time was 4.7days. The average hospital stay was 8.7 days. Operative complications were atelectasis(2), empyema with mediastinitis(1), recurrent laryngeal nerve palsy(1), and plenic nerve palsy(1). In conclusion, VATS for mediastinal lesions were performed with shorter operation time and hospital stay, and lesser complications and pain than those of conventional thoracotomy.

• Atmosphere
• /
• v.24 no.2
• /
• pp.173-188
• /
• 2014

The KMA has cooperated with the Oklahoma University in USA to develop a Polarimetric Radar Data (PRD) simulator to improve the microphysical processes in Korea Local Analysis and Prediction System (KLAPS), which is critical for the utilization of PRD into Numerical Weather Prediction (NWP) field. The simulator is like a tool to convert NWP data into PRD, so it enables us to compare NWP data with PRD directly. The simulator can simulate polarimetric radar variables such as reflectivity (Z), differential reflectivity ($Z_{DR}$), specific differential phase ($K_{DP}$), and cross-correlation coefficient (${\rho}_{hv}$) with input of the Drop Size Distribution (DSD) and scattering calculation of the hydrometeors. However, the simulator is being developed based on the foreign observation data, therefore the PRD simulator development reflecting rainfall characteristics of Korea is needed. This study analyzed a potential application of the 2-Dimension Video Disdrometer (2DVD) data by calculating the raindrop axis ratio according to the rain-types to reflect Korea's rainfall characteristics into scattering module in the simulator. The 2DVD instrument measures the precipitation DSD including the fall velocity and the shape of individual raindrops. We calculated raindrop axis ratio for stratiform, convective and mixed rainfall cases after checking the accuracy of 2DVD data, which usually represent the scattering characteristics of precipitation. The raindrop axis ratio obtained from 2DVD data are compared with those from foreign database in the simulator. The calculated the dual-polarimetric radar variables from the simulator using the obtained raindrop axis ratio are also compared with in situ dual-polarimetric observation data at Bislsan (BSL). 2DVD observation data show high accuracies in the range of 0.7~4.8% compared with in situ rain gauge data which represents 2DVD data are sufficient for the use to simulator. There are small differences of axis ratio in the diameter below 1~2 mm and above 4~5 mm, which are more obvious for bigger raindrops especially for a strong convective rainfall case. These differences of raindrop axis ratio between domestic and foreign rainfall data base suggest that the potential use of disdrometer observation can develop of a PRD simulated suitable to the Korea precipitation system.

• Journal of Korean Society of Environmental Engineers
• /
• v.39 no.5
• /
• pp.277-281
• /
• 2017

Lately rainfall characteristics that it rains a lot in a short space of time often occurs. Because of this meteorological phenomena, the flow rate and concentration of initial rainfall for runoff and combined sewer overflows are changed. In the case of this inlet fluctuation, the flotation method at high surface loading rate is suitable for water quality management. the flotation method is able to meet the removal rate requirements of water public zone in 5 to 10 min which is irelatively short period. For assessment and diagonision of flotation method, A/S ratio is applied until now. But unfortunately, this has some limits for evaluation standard for certification and assessment of technical diagnosis and operation. This is why there is different efficiency in the bubble distribution at the same A/S ratio. The velocity and time of floating is changed by the different bubble distributions. The floating time affects the plant volume because the time factor make size dicision. Therefore the charateristics of bubble distribution and floating time at the same A/S ratio is necessary to apply to evaluation standard for certification and assessment of technical diagnosis and operation. For generalization of the method in certification and assessment, the characteristics of bubble distribution was studied. Until recently, using the optical device and shooting live video, there are some analysis technology of the floating factors. But this kind of technology is influenced by the equipment. with this level of confidence about the results, it is difficult to apply to generalize. According this reasons, this study should be applied on experiment generalization of method about measurement of relation between bubble distribution and floating time.

• The Journal of the Korea Contents Association
• /
• v.9 no.4
• /
• pp.12-18
• /
• 2009

In this paper, we propose a novel method which efficiently codes $4{\times}4$ intra prediction modes in H.264/ AVC video encoding. There are nine optional prediction modes for each $4{\times}4$luma block and 4 modes for each $16{\times}16$luma block. To code each $4{\times}4$ intra prediction mode, lots of bits are required. To efficiently compress the intra mode bits in H.264/AVC, the most probable mode(MPM) is estimated by using the intra modes of the adjacent blocks, since intra modes for neighboring $4{\times}4$luma blocks are correlated. In this paper, a novel method for estimating the MPM is proposed by using the statistical distribution of intra modes of adjacent intra blocks. Experimental results show that the proposed method can achieve a coding gain of about 0.1dB.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.25 no.5
• /
• pp.301-310
• /
• 2013

For analyzing sand grain size, a specialized high-priced instrument has been used, such as sieve shaker, video camera, laser particle size analyzer, and microscope. Among these, the sieve shaker is commonly used because it is not only cheaper than others but also provides reasonable accuracy. However, it takes a long time and makes lots of dust and noise. In this study, a cheaper and easier method which can replace the sieve shaker is proposed. By using a commercial flat-bed scanner and a darkroom box, the sand size distribution can be analyzed. The darkroom box makes sand images clear and protects the glass of the scanner from being scratched. Comparison between the present method and sieve analysis shows that the present method not only has an accuracy comparable to the sieve analysis but also can save time and effort.

• Journal of Korea Multimedia Society
• /
• v.13 no.4
• /
• pp.582-593
• /
• 2010

Recently the commercial value of GIS(geographic information system) has been increased with growing application demands of spatial information in life space. Also, GIS has been expanded to the integrated information system that provides the ease access to map information by public users and also the combined service of GIS map data and various multimedia contents. But several security problems of GIS map data have brought up at GIS industries, such as the illegal copy and distribution of GIS map data. Furthermore, the copyright protection of GIS map data has not been researched yet compared with image and video contents. This paper presents a polyline based watermarking scheme for the copyright protection of GIS vector map. The proposed scheme analyzes the structure of vector map and embeds adaptively the watermark by using the distance distribution of polyline components. From experimental results, we verified that the proposed scheme satisfies the invisibility and the robustness of geometrical attacks.

• Journal of the Korean Society for Precision Engineering
• /
• v.24 no.7 s.196
• /
• pp.120-125
• /
• 2007

Analysis of elders' behavioral characteristic during showering provides important fundamental data when developing an automatic showering system. We recorded showering behavior of 37 elders' at the front, back, left and right view using four CCTV cameras. The photographic data were analyzed by counting the frequencies of body scrub and identifying zones where hands could reach. And we measured their body sizes, and conducted questionnaire of showering for 302 peoples. Analysis was done according to the analysis criteria established. The human body was divided into 14 zones from a body, and the numbers of body scrub for each zone were counted. The average total numbers of body scrub were 170.8 excluding the head, the face and the neck. The average numbers of body scrub with right hand were 70.2, left hand were 88.2, and both hands were 81.8. The percentages of body scrub frequencies were 16.3% (breast), 12.2%(crotch), 8.8% (abdomen), 7.5% (waist), 8.8% (shoulder), 7% (hips), 0.4% (back), 15.2% (arm), and 23.9% (leg). Elders scrubbed the breast, crotch and abdomen more frequently than the back and hips. According to the video analysis and questionnaire results, the distribution of shower heads (or nozzles) at the front side should be arranged at the ratio of 40% of the trunk(breast, shoulders, abdomen), 20% of the thighs, the calves, and the crotch, respectively. The distribution of shower heads (or nozzles) of the back side were arranged at the ratio of 20% of the waist and the arms, 20% of the back, the thighs, the calves, and the hip, respectively.

• Journal of Digital Convergence
• /
• v.13 no.4
• /
• pp.377-385
• /
• 2015

The purpose of the study was to perform trend analysis of the exercise content i.e. element difficulty distribution scores, element group distribution scores obtained by the athletes during 2014 Korea Cup International Gymnastic finals. The conclusion drawn from the analysis are as follows: Firstly in terms of tournament technical difficulty value, Korean athletes showed characteristics of reliance on swing element rather than strength hold element which could actually secure higher difficulty scores. Secondly, skill acquisition of higher difficulty value is demanded by taking advantage of familiar characteristics from the swing element i.e. Jonasson and Roll bwd. Slowly with str. arms and body to swallow (2s.). Thirdly, development of sensory training and strengthening exercise program are essential to enhance strength hold technique.