• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2014.11a
• /
• pp.268-270
• /
• 2014

Object extraction problem is one of the most important topics in the research area of computer vision, this type of technique can be widely used in practical, such as image processing, robot vision, automatically traffic guide and so on. In this paper, we propose a different way to estimate the background and foreground without any previous training procedure, this approach can be used for automatic object extraction in the future. A simple experiment result shows that our approach has a good potential for the further more practical application.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2020.11a
• /
• pp.244-247
• /
• 2020

Data augmentation has been helpful in improving the performance in deep learning, when we have a limited data and random erasing is one of the augmentations that have shown impressive performance in deep learning in multiple domains. But the main issue is that sometime it loses good features when randomly selected region is erased by some random values, that does not improve performance as it should. We target that problem in way that good features should not be lost and also want random erasing at the same time. For that purpose, we introduce new augmentation technique named Intra-Class Random Erasing (ICRE) that focuses on data to learn robust features of the same class samples by randomly exchanging randomly selected region. We perform multiple experiments by using different models including resnet18, VGG16 over variety of the datasets including ESC10, UrbanSound8K. Our approach has shown effectiveness over others methods including random erasing.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2022.06a
• /
• pp.182-183
• /
• 2022

The large data volume of light field (LF) image has motivated much research on how to compress the data volume more efficiently. One of the approaches is to compress LF images after representing them in the form of pseudo video sequence. In this way, the pseudo temporal redundancy between views can be exploited by motion estimation and compensation. Based on our observation that images obtained by LF cameras have small range of disparity values between adjacent views, we propose to limit the motion search range to reduce the time complexity of motion estimation. Our experimental results show that a smaller motion search range reduces the encoding time while not affecting the bitrate of H.266/VVC much.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.20 no.5
• /
• pp.957-962
• /
• 2016

Current terrestrial Digital Multimedia Broadcasting (T-DMB) is conducting the emergency alert broadcast, or can view a variety of broadcasting. However, propagation shadow area is a situation where the service is limited due to limitations of facilities investment. In addition, there is the problem of T-DMB broadcasting is for viewing only a restricted area and a mobile device because the mobile is also T-DMB viewing device impossible. In this paper, it receives a T-DMB broadcasting as a way to solve the problems of the T-DMB system, which was studied the re transmission to the mobile device. Accordingly, by receiving the broadcast may be watched in the mobile device the T-DMB reception impossible. Also provides a one-way/two-way authentication mechanism using a conditional access function, and the system was configured so that the user can watch only the registered broadcasting.

• Journal of Broadcast Engineering
• /
• v.20 no.4
• /
• pp.598-605
• /
• 2015

Effective alerts will drastically mitigate the disaster impacts. One way to enhance the effectiveness is to prepare multiple alert distribution channels. In this paper, we propose a new emergency alert distribution system that will solve the weaknesses of the current distribution systems, such as Cell Broadcasting System and T-DMB Automatic Emergency Alert Service. The weaknesses are: the limitation of message length, small portions of terminals that can receive and display the alert signal, and one-way communication service. To solve these problems, we propose an emergency alert distribution system over the Internet with RSS (Rich Site Summary) format. The system also has the capability of adaptive alert filtering according to the receiver characteristics. We analyzed the characteristics of RSS as an alerting format, and draw functional requirements satisfying use case scenarios. We designed the system only with major requirements and verified it on our test bed. Sending richer message contents through various receiving terminals, the system will achieve more effective emergency alert service.

• The KIPS Transactions:PartC
• /
• v.17C no.6
• /
• pp.499-504
• /
• 2010

The flooding is the simplest and effective way to disseminate a packet to all nodes in a wireless sensor network (WSN). However, basic flooding makes all nodes transmit the packet at least once, resulting in the broadcast storm problem in a serious case, in turn network resources become severely wasted. Particularly, power is one of the most valuable resources of WSNs as nodes are powered by battery, then the waste of energy by the basic flooding lessens the lifetime of WSNs. In order to solve the broadcast storm problem, this paper proposes a dynamic probabilistic flooding that utilizes the neighbor information like the number of child and sibling nodes. Simulation results show that the proposed method achieves a higher packet delivery ratio with the similar number of duplicate packets as compared to existing schemes.

• Journal of Broadcast Engineering
• /
• v.11 no.1 s.30
• /
• pp.54-65
• /
• 2006

In this paper, we propose a new functionality to Scalable Video Coding (SVC), which is the support of multiple ROIs for heterogeneous display resolution. Scalable video coding is targeted at giving temporal, spatial, and quality scalability for the encoded bit stream. Region of interest (ROI) is an area that is semantically important to a particular user, especially users with heterogeneous display resolutions. The bitstream containing the ROIs could to be extracted without any transcoding operations, which may be one of way to satisfy QoS. To define multiple ROI in SVC, we adapted FMO, a tool defined in H.264, and based on it, we propose a way to encode and decode ROIs. The proposed method is implemented on the JSVM1.0 and the functionality is verified using it.

• Proceedings of the IEEK Conference
• /
• 2007.07a
• /
• pp.49-50
• /
• 2007

We are now using separate ID Database, Permission System, OS and Standard Data Format in PC, Cell phone, IPTV service and these are not good in many kind of usage, data sharing, contents servicing and supervising efficiency To merge these multiple services into one system we need to make a single ID systems for all and this paper will gide you to the new ID system that was much better than the old one. This new type of ID system is made with 3D Data and this will give fantastic visual effect & rich information data to the ID users in 3D way which is much better than 2D image & Text information technology.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 1999.06a
• /
• pp.83-88
• /
• 1999

In this paper, a novel gray-scale lossless image coder combining context-based minimum mean squared error (MMSE) prediction and entropy coding is proposed. To obtain context of prediction, this paper first defines directional difference according to sharpness of edge and gradients of localities of image data. Classification of 4 directional differences forms“geometry context”model which characterizes two-dimensional general image behaviors such as directional edge region, smooth region or texture. Based on this context model, adaptive DPCM prediction coefficients are calculated in MMSE sense and the prediction is performed. The MMSE method on context-by-context basis is more in accord with minimum entropy condition, which is one of the major objectives of the predictive coding. In entropy coding stage, context modeling method also gives useful performance. To reduce the statistical redundancy of the residual image, many contexts are preset to take full advantage of conditional probability in entropy coding and merged into small number of context in efficient way for complexity reduction. The proposed lossless coding scheme slightly outperforms the CALIC, which is the state-of-the-art, in compression ratio.

• Journal of Korean Society of Forest Science
• /
• v.31 no.1
• /
• pp.37-42
• /
• 1976

In this study, an effort was made to find out more effective methods for vegetation establishment on denuded forest land along Cheon Bo Mountains, This area has serious sheet erosion as a major erosion process and parent material is granite gneiss which is known as one of the most erosive materials. Lespedeza bicolor, Arundinella hirta and Robinia pseudoacacia were used as ground cover species. Seeding methods were open broadcast seeding and broadcast seeding with straw mulch, dill seeding with 10cm and 20cm width, and spot seeding with 20cm and 30cm diameter. 1. On slopes less than 30 degrees, broadcast seeding method with straw mulch was the most effective way to establish vegetation cover. 2. On steep slopes more than 30 degrees, 10cm drill seeding method was more effective than the other methods. 3. Soil Texture has an influence on vegetation establishment governing soil moisture condition. Coverage and fresh weight were increased with the fine material qantity in soil. 4. In process of year, coverage was increased with fresh weight. On the other hand, survival decreased.