• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.41 no.6
• /
• pp.253-260
• /
• 2004

This paper proposes a novel algorithm for hybrid lossless audio coding, which employs an integer wavelet transform and a linear prediction model. The proposed algorithm divides the input signal into flames of a proper length, decorrelates the framed data using the integer wavelet transform and linear prediction and finally entropy-codes the frame data. In particular, the adaptive Golomb-Rice coding method used for the entropy coding selects an optimal option which gives the best compression efficiency. Since the proposed algorithm uses integer operations, it significantly improves the computation speed in comparison with an algorithm using real or floating-point operations. When the coding algorithm is implemented in hardware, the system complexity as well as the power consumption is remarkably reduced. Finally, because each frame is independently coded and is byte-aligned with respect to the frame header, it is convenient to move, search, and edit the coded, compressed data.

• Journal of the Korea Society of Computer and Information
• /
• v.16 no.2
• /
• pp.71-81
• /
• 2011

In order to easily create video tutorials, the algorithm is needed that video camera recording, white board images, video attachments, and document data are combined in real-time. In this study, we implemented the video lecture content creation system using wavelet-based lossless compression to composite multimedia objects in real-time and reproduce the images. Using commercially available PC can be useful when lecturers want to make video institutional contents, it can be operated easily and fastly. Therefore, it can be very efficient system for e-Learning and m-Learning. In addition, the proposed system including multimedia synthesis technology and real-time lossless compression technology can be applied to various fields, different kinds of multimedia creation, remote conferencing, and e-commerce so there are highly significant.

• Journal of Broadcast Engineering
• /
• v.18 no.2
• /
• pp.185-195
• /
• 2013

In this paper, we propose a video summarization technique on a smartphone, based on spatiotemporal saliency. The proposed technique detects scene changes by computing the difference of the color histogram, which is robust to camera and object motion. Then the similarity between adjacent frames, face region, and frame saliency are computed to analyze the spatiotemporal saliency in a video clip. Over-segmented hierarchical tree is created using scene changes and is updated iteratively using mergence and maintenance energies computed during the analysis procedure. In the updated hierarchical tree, segmented frames are extracted by applying a greedy algorithm on the node with high saliency when it satisfies the reduction ratio and the minimum interval requested by the user. Experimental result shows that the proposed method summaries a 2 minute-length video in about 10 seconds on a commercial smartphone. The summarization quality is superior to the commercial video editing software, Muvee.

• Journal of KIISE
• /
• v.45 no.3
• /
• pp.311-320
• /
• 2018

Recently, the use of digital audio and video as proof in criminal and all kinds of litigation is increasing, and scientific investigation using digital forensic technique is developing. With the development of computing and file editing technologies, anyone can simply manipulate video files, and the number of cases of manipulating digital data is increasing. As a result, the integrity of the evidence and the reliability of the evidence Is required. In this paper, we propose a technique for extracting the Electrical Network Frequency (ENF) through a grid of power grids according to the geographical environment for power supply, and then performing signal processing for peak detection using QIFFT. Through the detection algorithm using the standard deviation, it was confirmed that the video file was falsified with 73% accuracy and the forgery point was found.

• Journal of Broadcast Engineering
• /
• v.18 no.5
• /
• pp.680-690
• /
• 2013

In this paper, we propose a real-time temporal synchronization and compensation algorithm in stereoscopic video. Many temporal asynchronies are caused in the video editing stage and due to different transmission delays. These temporal asynchronies can degrade the perceived 3D quality. The goal of temporal alignment is to detect and to measure the temporal asynchrony and recover synchronization of the two video streams. In order to recover synchronization of the two video streams, we developed a method to detect asynchronies between the left and the right video streams based on a novel spatiogram information, which is a richer representation, capturing not only the values of the pixels but their spatial relationships as well. The proposed novel spatiogram additionally includes the changes of the spatial color distribution. Furthermore, we propose a block-based method for detection of the pair frame instead of one frame-based method. Various 3D experiments demonstrate the effectiveness of the proposed method.

• Journal of the Korea Society of Computer and Information
• /
• v.13 no.6
• /
• pp.87-94
• /
• 2008

Iris diagnosis is an alternative medicine technique whose proponents believe that patterns, colors, and other characteristics of the iris can be examined to determine information about a patient's systemic health. However, because most of previous studies only find a sign pattern in a gray iris image, they are not enough to be used for a iris diagnosis system. In this paper, we developed an iris diagnosis system based on a color image processing and iris medical information stored in database. The system includes four modules : input module with an iris camera, iris signs extraction module, medical database, output module with printing. Based on a color image processing approach, this paper presents the extraction algorithms of 7 major iris signs and correction manually for improving the accuracy of analysis. We can use the iridology and patient's health DB in the stage of signs analysis. Compared with the existing system, the developed system can be applied to an iris diagnosis system since it provides various additional functions.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2011.06a
• /
• pp.34-36
• /
• 2011

ENCs is edited based on the numerical charts for publishing paper charts and serviced in forms of grid styles. For this reason, the density of sounding information of ENCs is not consistent and was required for improvement. In this study, K-Means, ISODATA clustering algorithm as classification methods for satellite image was reviewed and adopted to case study. The developed results include loading module of ENC data, improvement algorithm of sounding information, writing module of ENC data. According to the results of algorithm, we could confirm the improved result.

• Journal of Korean Society of Transportation
• /
• v.20 no.7
• /
• pp.69-76
• /
• 2002

In conducting studies related to the national base networks, it is very important to construct a simulation network. This research provides an algorithm to construct the simulation network from the digital transportation map which is constructed based on the National Geographic Information System(NGIS). The algorithm consists of three functions(extraction of networks from transportation digital map, transform of the derived network into one suitable to transportation simulation model, and inspection of errors in the network). The direct derivation of a simulation network from GIS enables to enhance the reliability of an analysis related to a transportation facility investment as well as to reduce cost and time. In this research, Emme/2 which is generally accepted transportation planning fields is adapted for the target system. However, this algorithm will be extended to other simulation models such as Satong-Paldal which is the only transportation simulation model developed in Korea, and Tranplan in the near future.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.10
• /
• pp.1-7
• /
• 2021

Recently, since most of the research on correcting speech recognition errors is based on English, there is not enough research on Korean speech recognition. Compared to English speech recognition, however, Korean speech recognition has many errors due to the linguistic characteristics of Korean language, such as Korean Fortis and Korean Liaison, thus research on Korean speech recognition is needed. Furthermore, earlier works primarily focused on editorial distance algorithms and syllable restoration rules, making it difficult to correct the error types of Korean Fortis and Korean Liaison. In this paper, we propose a context-sensitive post-processing model of speech recognition using a LSTM-based sequence-to-sequence model and Bahdanau attention mechanism to correct Korean speech recognition errors caused by the pronunciation. Experiments showed that by using the model, the speech recognition performance was improved from 64% to 77% for Fortis, 74% to 90% for Liaison, and from 69% to 84% for average recognition than before. Based on the results, it seems possible to apply the proposed model to real-world applications based on speech recognition.

• KIPS Transactions on Software and Data Engineering
• /
• v.7 no.11
• /
• pp.411-418
• /
• 2018

With the proliferation of artificial intelligence technology, it is becoming important to obtain, store, and utilize scientific data in research and science sectors. A number of methods for extracting meaningful objects such as graphs and tables from research articles have been proposed to eventually obtain scientific data. Existing extraction methods using heuristic approaches are hardly applicable to electronic documents having heterogeneous manuscript formats because they are designed to work properly for some targeted manuscripts. This paper proposes a prototype of an object extraction system which exploits a recent deep-learning technology so as to overcome the inflexibility of the heuristic approaches. We implemented our trained model, based on the Faster R-CNN algorithm, using the Google TensorFlow Object Detection API and also composed an annotated data set from 100 research articles for training and evaluation. Finally, a performance evaluation shows that the proposed system outperforms a comparator adopting heuristic approaches by 5.2%.