• The Journal of the Korea Contents Association
• /
• v.8 no.3
• /
• pp.8-17
• /
• 2008

A new algorithm for automatic extraction of interesting objects is proposed in this paper. The proposed algorithm can be summarized in two steps. First, segmentation of color image that split interesting objects and backgrounds is performed. According to the research stating, 'Humans perceive things by contracting color into three to four essential colors,' a color image is segmented into three regions utilizing k-mean algorithm, followed by annexing the regions when the similarities of them exceeds the critical value based on the calculation of degrees in the histogram similarity, Second, identifying the interesting objects out of the segmented image, partitioned by the image composition theory, is performed. To have a good picture, it is important to adjust positions of interesting objects according to picture composition. Extracting objects is a retro-deduction process using a weighted mask designed upon the triangular composition of picture. To prove the quality of the proposed method, experiments are performed over four hundreds images as well as comparison with recently proposed KMCC and GBIS methods.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.8
• /
• pp.3295-3311
• /
• 2020

With the wide application of hyperspectral images, it becomes more and more important to compress hyperspectral images. Conventional recursive least squares (CRLS) algorithm has great potentiality in lossless compression for hyperspectral images. The prediction accuracy of CRLS is closely related to the correlations between the reference bands and the current band, and the similarity between pixels in prediction context. According to this characteristic, we present an improved CRLS with adaptive band selection and adaptive predictor selection (CRLS-ABS-APS). Firstly, a spectral vector correlation coefficient-based k-means clustering algorithm is employed to generate clustering map. Afterwards, an adaptive band selection strategy based on inter-spectral correlation coefficient is adopted to select the reference bands for each band. Then, an adaptive predictor selection strategy based on clustering map is adopted to select the optimal CRLS predictor for each pixel. In addition, a double snake scan mode is used to further improve the similarity of prediction context, and a recursive average estimation method is used to accelerate the local average calculation. Finally, the prediction residuals are entropy encoded by arithmetic encoder. Experiments on the Airborne Visible Infrared Imaging Spectrometer (AVIRIS) 2006 data set show that the CRLS-ABS-APS achieves average bit rates of 3.28 bpp, 5.55 bpp and 2.39 bpp on the three subsets, respectively. The results indicate that the CRLS-ABS-APS effectively improves the compression effect with lower computation complexity, and outperforms to the current state-of-the-art methods.

• Geophysics and Geophysical Exploration
• /
• v.12 no.2
• /
• pp.208-214
• /
• 2009

Using the GNSS data and tilt-meter data of Boeun (BOEN) and Goesan (GSAN) GNSS stations, we have calculated the differential distance vector variation with the calculation time span set to 1 hour and 3 hour and differential tilt vector variation along time and derived an indicator of similarity between the two variations along time. The similarity such calculated is rather lower than high. But as the existence of a circular type movement of the antenna's phase center's location due to the tilt's variation of the antenna tower because of the sunlight's diurnal change is certain, we recommend to take such diurnal variation of antenna's location into consideration when the correction error in DGNSS or the measured data at reference stations in VRS (Virtual Reference System) is broadcast.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.27 no.1
• /
• pp.87-96
• /
• 2023

As the number of security vulnerabilities increases yearly, security threats continue to occur, and the vulnerability risk is also important. We devise a security threat score calculation reflecting trends to determine the risk of security vulnerabilities. The three stages considered key elements such as attack type, supplier, vulnerability trend, and current attack methods and techniques. First, it reflects the results of checking the relevance of the attack type, supplier, and CVE. Secondly, it considers the characteristics of the topic group and CVE identified through the LDA algorithm by the Jaccard similarity technique. Third, the latest version of the MITER ATT&CK framework attack method, technology trend, and relevance between CVE are considered. We used the data within overseas sites provide reliable security information to review the usability of the proposed final formula CTRS. The scoring formula makes it possible to fast patch and respond to related information by identifying vulnerabilities with high relevance and risk only with some particular phrase.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.11
• /
• pp.4084-4104
• /
• 2021

Visual place recognition (VPR) is a fundamental yet challenging task of mobile robot navigation and localization. The existing VPR methods are usually based on some pairwise similarity of image descriptors, so they are sensitive to visual appearance change and also computationally expensive. This paper proposes a simple yet effective four-step method that achieves adaptive convolutional neural network (CNN) landmark matching for VPR. First, based on the features extracted from existing CNN models, the regions with higher significance scores are selected as landmarks. Then, according to the coordinate positions of potential landmarks, landmark matching is improved by removing mismatched landmark pairs. Finally, considering the significance scores obtained in the first step, robust image retrieval is performed based on adaptive landmark matching, and it gives more weight to the landmark matching pairs with higher significance scores. To verify the efficiency and robustness of the proposed method, evaluations are conducted on standard benchmark datasets. The experimental results indicate that the proposed method reduces the feature representation space of place images by more than 75% with negligible loss in recognition precision. Also, it achieves a fast matching speed in similarity calculation, satisfying the real-time requirement.

• The Journal of Bigdata
• /
• v.6 no.2
• /
• pp.51-60
• /
• 2021

The market size of wearable devices is growing rapidly every year, and manufacturers around the world are introducing products that utilize their unique characteristics to keep up with the demand. Among them, smart watches are wearable devices with a very high share in sales, and they provide a variety of services to users by using information collected in real-time. The quality of service depends on the accuracy of the data collected by the smart watch, but data measurement may not be possible depending on the situation. This paper introduces a method to restore data that a smart watch could not collect. It deals with the similarity calculation method of trajectory information measured over time for data restoration and introduces a procedure for restoring missing sections according to the similarity. To prove the performance of the proposed methodology, a comparative experiment with a machine learning algorithm was conducted. Finally, the expected effects of this study and future research directions are discussed.

• Cartoon and Animation Studies
• /
• s.40
• /
• pp.477-496
• /
• 2015

Emotional awareness is the image of a person is represented by different tendencies. Currently, the emotion computing to objectively evaluate the emotion recognition research is being actively studied. However, existing emotional computing research has many problems to run. First, the non-objective in emotion recognition if it is inaccurate. Second, the correlation between the emotion recognition is unclear points. So to test the regularity of image sensitivity to the need of the present study is to control emotions in the computing system. In addition, the screen number of the emotion recognized for the purpose of this study, applying the method of objective image emotional computing system and compared with a similar degree of emotion of the person. The key features of the image emotional computing system calculates the emotion recognized as numbered digital form. And to study the background of emotion computing is a key advantage of the effect of the James A. Russell for digitization of emotion (Core Affect). Pleasure emotions about the core axis (X axis) of pleasure and displeasure, tension (Y-axis) axis of tension and relaxation of emotion, emotion is applied to the computing research. Emotional axis with associated representative sensibility very happy, excited, elated, happy, contentment, calm, relaxing, quiet, tired, helpless, depressed, sad, angry, stress, anxiety, pieces 16 of tense emotional separated by a sensibility ComputingIt applies. Course of the present study is to use the color of the color key elements of the image computing formula sensitivity, brightness, and saturation applied to the sensitivity property elements. Property and calculating the rate sensitivity factors are applied to the importance weight, measured by free-level sensitivity score (X-axis) and the tension (Y-axis). Emotion won again expanded on the basis of emotion crossed point, and included a representative selection in Sensibility size of the top five ranking representative of the main emotion. In addition, measuring the emotional image of a person with 16 representative emotional score, and separated by a representative of the top five senses. Compare the main representative of the main representatives of Emotion and Sensibility people aware of the sensitivity of the results to verify the similarity degree computing emotion emotional emotions depending on the number of representative matches. The emotional similarity computing results represent the average concordance rate of major sensitivity was 51%, representing 2.5 sensibilities were consistent with the person's emotion recognition. Similar measures were the degree of emotion computing calculation and emotion recognition in this study who were given the objective criteria of the sensitivity calculation. Future research will need to be maintained weight room and the study of the emotional equation of a higher concordance rate improved.

• Journal of KIISE
• /
• v.44 no.10
• /
• pp.1087-1093
• /
• 2017

Educating foreigners in Korean language is attracting increasing attention with the growing number of foreigners who want to learn Korean or want to reside in Korea. Existing spell checkers mostly focus on native Korean speakers, so they are inappropriate for foreigners. In this paper, we propose a correction method for the Korean language that reflects the contextual characteristics of Korean and writing characteristics of foreigners. Our method can extract frequently used expressions by Koreans by constructing syllable reverse-index for eojeol bi-gram extracted from corpus as correction candidates, and generate ranked Korean corrections for foreigners with upgraded edit distance calculation. Our system provides a user interface based on keyboard hooking, so a user can easily use the correction system along with other applications. Our system improves the detection rate for foreign language users by about 45% compared to other systems in foreign language writing environments. This will help foreign users to judge and correct their own writing errors.

• Nuclear Engineering and Technology
• /
• v.40 no.3
• /
• pp.199-212
• /
• 2008

A small-break loss of coolant accident (SB-LOCA) test with a break size equivalent to a 3-inch cold leg break of the APR1400 was carried out as the first transient integral effect test using the ATLAS (Advanced Thermal-hydraulic Test Loop for Accident Simulation). This was the first integral effect test to investigate the integral performance of the test facility and to verify its simulation capability for one of the design-basis accidents. Reasonably good thermal hydraulic data was obtained so that an integral performance of the fluid sub-systems was identified and control performance of the ATLAS was confirmed under real thermal hydraulic conditions. Based on the measured data, a post-test calculation was carried out using the best-estimate thermal hydraulic safety analysis code, MARS 3.1, and the similarity between the expected and actual data was investigated. On the whole, the post-test calculation reasonably predicts the major thermal hydraulic parameters measured during the SB-LOCA test. The obtained data will be used to enhance the simulation capability of the ATLAS and to improve an input model of the ATLAS for simulation of other target scenarios.

• Journal of KIISE:Software and Applications
• /
• v.27 no.7
• /
• pp.769-783
• /
• 2000

In this paper, in order to disambiguate Korean noun word sense, we define a local context and explain how to extract it from a raw corpus. Following the intuition that two different nouns are likely to have similar meanings if they occur in the same local context, we use, as a clue, the word that occurs in the same local context where the target noun occurs. This method increases the usability of extracted knowledge and makes it possible to disambiguate the sense of infrequent words. And we can overcome the data sparseness problem by extending the verbs in a local context. The sense of a target noun is decided by the maximum similarity to the clues learned previously. The similarity between two words is computed by their concept distance in the sense hierarchy borrowed from WordNet. By reducing the multiplicity of clues gradually in the process of computing maximum similarity, we can speed up for next time calculation. When a target noun has more than two local contexts, we assign a weight according to the type of each local context to implement the differences according to the strength of semantic restriction of local contexts. As another knowledge source, we get a co-occurrence information from dictionary definitions and example sentences about the target noun. This is used to support local contexts and helps to select the most appropriate sense of the target noun. Through experiments using the proposed method, we discovered that the applicability of local contexts is very high and the co-occurrence information can supplement the local context for the precision. In spite of the high multiplicity of the target nouns used in our experiments, we can achieve higher performance (89.8%) than the supervised methods which use a sense-tagged corpus.