• The Sea:JOURNAL OF THE KOREAN SOCIETY OF OCEANOGRAPHY
• /
• v.4 no.4
• /
• pp.323-337
• /
• 1999

Seafloor morphology and manganese nodule occurrence were studied in the Korea Deep-sea Environmental Study (KODES) area, northeast equatorial Pacific, to understand their relationship. Study area is composed of three elongated valleys and hills with about 100~200 m height along NNE-SSW direction. Valley region is generally flat. However, hill region is very rugged with big cliffs of about 100m height and small depressions of several tens of meters depth. Tectonic movement along the Clarion-Clipperton fracture zone, consequent formation of elongated abyssal hills and Valleys, erosion of siliceous bottom sediments by bottom currents, and dissolution of carbonate sediments on the abyssal hills below CCD result in the rugged morphology. Manganese nodule occurrence is closely related to the morphology of the study area; mostly rounded-shaped manganese nodules with about 5 cm diameter are abundant on the flat valley region, whereas irregular shaped nodules (or manganese crust) with less than 5 cm to about 1 m diameter occur on the hill. These results supports the previous reports that nodule abundance, composition, and morphology are variable both on regional and local small scales on the seafloor even within some abundant nodule provinces depending on oceanographic characteristics such as bathymetric features, surface sediment type, sediment thickness, and so on. We suggest that such oceanographic characteristics affect interrelatedly on the formation of manganese nodules, and tectonic movement of the Pacific plate ultimately constrain the nodule occurrence. A potential mining place in the KODES area seems to be the valley region, which is elongated to the NNW-SSE direction with 3-4 km width.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.14 no.1
• /
• pp.143-150
• /
• 1994

When the transverse bed slope ($S_t$) in channel bend is more than 0.1, it may produce undesirable results on the bed topography of the cross section. The linear relationship for $S_t$ results in zero or negative flow depths at the shallow $S_t$de of the cross section (i.e., inner bank). The exponential relationship for $S_t$ results in excessive flow depths at the deep side of the cross section (i.e., outer bank). This problem can be solved by combining the best features of both relationships described above. From the study, the linear relationship can be applied for the deep $S_t$de of the cross section. But the exponential relationship is suitable for the shallow side. Therefore, the new relationship of $S_t$ is clarified mathematically. A new mathematical model for bed topography is developed herein which takes accounts of the phase lag and the influence of the width to depth ratio. This model is used to analyze two sets of data: one from laboratory channel and the other from natural channel. A good agreement is found between the observed and the calculated bed topography based on the analysis of two sets of data.

• Journal of Yeungnam Medical Science
• /
• v.33 no.2
• /
• pp.166-169
• /
• 2016

Afferent loop obstruction following gastrectomy is a rare but fatal complication. Clinical features of afferent loop obstruction are mainly gastrointestinal symptoms. A 56-year-old female underwent radical total gastrectomy with Roux-en-Y esophagojejunostomy for treatment of advanced gastric cancer. After fourteen months postoperatively, she showed gradual development of edema of both legs. Computed tomography (CT) scan showed disease progression at the jejunojejunostomy site and consequent dilated afferent loop, which resulted in inferior vena cava (IVC) compression. A drainage catheter was placed percutaneously into the afferent loop through the intrahepatic duct and an IVC filter was placed at the suprarenal IVC, and self-expanding metal stents were inserted into bilateral common iliac veins. With these procedures, sympotms related with afferent loop obstruction and deep vein thrombosis were improved dramatically. The follow-up abdominal CT scan was taken 3 weeks later and revealed the completely decompressed afferent loop and improved IVC patency. Surgical treatment should be considered as the first choice for afferent loop obstruction; however, because it is more immediate and less invasive, non-surgical modalities, such as percutaneous catheter drainage or stent placement, can be effective alternatives for inoperable cases or risky patients who have severe medical comorbidities.

• Journal of the Korea Convergence Society
• /
• v.9 no.8
• /
• pp.1-8
• /
• 2018

In the field of biometric security such as face and iris recognition, it is essential to extract facial features such as eyes and lips. In this paper, we have studied a method of detecting eye and lip region in face image using faster R-CNN. The faster R-CNN is an object detection method using deep running and is well known to have superior performance compared to the conventional feature-based method. In this paper, feature maps are extracted by applying convolution, linear rectification process, and max pooling process to facial images in order. The RPN(region proposal network) is learned using the feature map to detect the region proposal. Then, eye and lip detector are learned by using the region proposal and feature map. In order to examine the performance of the proposed method, we experimented with 800 face images of Korean men and women. We used 480 images for the learning phase and 320 images for the test one. Computer simulation showed that the average precision of eye and lip region detection for 50 epoch cases is 97.7% and 91.0%, respectively.

• Korean Journal of Soil Science and Fertilizer
• /
• v.43 no.5
• /
• pp.565-571
• /
• 2010

One of the disadvantages of flooding treatment for desalting from soils is that salts move to deep soils after flooding and at the end reaccumulate at the soil surface through capillary movements. This study was carried out to remove salts from soils in plastic film houses by a modified flooding method, drainage after flooding. The method successfully removed salts at the soil surface and salts did not move to the deep soil. Drained water containing N, P and K could be reused as fertilizer. By applying small amount of MgO, turbidity of water flooded decreased in 30 min by 95%. Struvite should be formed since the flooded water contain ammonia and phosphorous and their concentrations were decreased. This could be utilized as fertilizer which provides a slow-release source of phosphorus, magnesium and nitrogen that features low inherent water solubility.

• The Korean Journal of Community Living Science
• /
• v.18 no.1
• /
• pp.107-129
• /
• 2007

The purpose of this research is to understand features of the present condition of green tea food by analyzing the data on tea foods presented in Korean literature after the 1990's (two articles from professional journals related to tea culture, and three books related to tea food). The main conclusions are as follows: 1. It is found from separating 354 different kinds of green tea foods into three categories - main dishes, side dishes and desserts - that there are 137 kinds of side dishes, 123 kinds of desserts, and 94 kinds of main dishes from green tea foods. Upon dividing these into smaller categories, there are 40 rice dishes, 27 noodle dishes, 18 gruel dishes and 9 dumpling dishes found among the main dishes; 26 pan fried dishes, 24 potherb/cooked potherbs dishes, 17 deep-fried dishes, 15 soup/broth dishes, 14 grilled dishes, 11 smothered dishes, 10 hard -boiled/fried dishes, 6 kimchi dishes, 4 dried food dishes, 4 jelly dishes, 4 stew dishes, and 2 raw fish dishes among the side dishes; and 37 snack dishes, 36 punch/drink dishes, 26 rice cake dishes, and 24 bread dishes are found among the desserts. 2. There are 201 kinds of green tea foods using powders, 107 kinds using wet tea leaves, 61 kinds using dry tea leaves, 57 kinds using water of drawn tea, and 17 kinds using wild tea leaves, according to analysis of teas used for green tea foods. There is more use of powder for snacks, punch and drinks, rice cakes, noodles, and breads, and more use of wet tea leaves for rice, pan fried food, and potherb/cooked potherb dishes. It is also shown that there is more use of water from drawn tea for rice, punch and drinks, noodles, and gruels, more use of dry tea leaves for snack, rice, breads, and more use of wild tea leaves for deep-fried and pan fried kinds of tea foods.

• Journal of Korea Multimedia Society
• /
• v.22 no.12
• /
• pp.1415-1429
• /
• 2019

Nowadays object tracking process becoming one of the most challenging task in Computer Vision filed. A CSR-DCF (channel spatial reliability-discriminative correlation filter) tracking algorithm have been proposed on recent tracking benchmark that could achieve stat-of-the-art performance where channel spatial reliability concepts to DCF tracking and provide a novel learning algorithm for its efficient and seamless integration in the filter update and the tracking process with only two simple standard features, HoGs and Color names. However, there are some cases where this method cannot track properly, like overlapping, occlusions, motion blur, changing appearance, environmental variations and so on. To overcome that kind of complications a new modified version of CSR-DCF algorithm has been proposed by integrating deep learning based object detection and CSRT tracker which implemented in OpenCV library. As an object detection model, according to the comparable result of object detection methods and by reason of high efficiency and celerity of Faster RCNN (Region-based Convolutional Neural Network) has been used, and combined with CSRT tracker, which demonstrated outstanding real-time detection and tracking performance. The results indicate that the trained object detection model integration with tracking algorithm gives better outcomes rather than using tracking algorithm or filter itself.

• Journal of the Korea Society of Computer and Information
• /
• v.24 no.12
• /
• pp.9-16
• /
• 2019

The goal of this study is to propose an efficient model for recognizing and classifying tree images to measure the accuracy that can be applied to smart devices during class. From the 2009 revised textbook to the 2015 revised textbook, the learning objective to the fourth-grade science textbook of elementary schools was added to the plant recognition utilizing smart devices. In this study, we compared the recognition rates of trees before and after retraining using a pre-trained inception V3 model, which is the support of the Google Inception V3. In terms of tree recognition, it can distinguish several features, including shapes, bark, leaves, flowers, and fruits that may lead to the recognition rate. Furthermore, if all the leaves of trees may fall during winter, it may challenge to identify the type of tree, as only the bark of the tree will remain some leaves. Therefore, the effective tree classification model is presented through the combination of the images by tree type and the method of combining the model for the accuracy of each tree type. I hope that this model will apply to smart devices used in educational settings.

• Journal of Broadcast Engineering
• /
• v.25 no.4
• /
• pp.569-575
• /
• 2020

The serial arc is one of factors causing electrical fires. Over past decades, various researches have been carried out to detect arc occurrences. Even though frequency analysis, wavelet, and statistical features have been used, additional steps such as transformation and feature extraction are required. On the contrary, deep learning models directly use the raw data without any feature extraction processes. Therefore, the usage of time-domain data is preferred, but the performance is not satisfactory. To solve this problem, subsequent 1-D signals are transformed into 2-D data that can feed into a convolutional neural network (CNN). Experiments validated that CNN model outperforms deep neural network (DNN) by the classification accuracy of 8.6%. In addition, data augmentation is utilized, resulting in the accuracy improvement by 14%.

• Journal of Broadcast Engineering
• /
• v.22 no.2
• /
• pp.162-172
• /
• 2017

In this paper, we propose facial expression recognition using CNN (Convolutional Neural Network), one of the deep learning technologies. To overcome the disadvantages of existing facial expression databases, various databases are used. In the proposed technique, we construct six facial expression data sets such as 'expressionless', 'happiness', 'sadness', 'angry', 'surprise', and 'disgust'. Pre-processing and data augmentation techniques are also applied to improve efficient learning and classification performance. In the existing CNN structure, the optimal CNN structure that best expresses the features of six facial expressions is found by adjusting the number of feature maps of the convolutional layer and the number of fully-connected layer nodes. Experimental results show that the proposed scheme achieves the highest classification performance of 96.88% while it takes the least time to pass through the CNN structure compared to other models.