• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.12 no.4
• /
• pp.707-715
• /
• 2008

This paper proposes an algorithm for efficient compression gray-level images by entropy encoder. The issue of the proposed method is to replace original data of gray-level images with particular ranked data. For this, first, before encoding a stream of gray-level values in an image, the proposed method counts co-occurrence frequencies for neighboring pixel values. Then, it replaces each pay value with particularly ranked numbers based on the investigated co-occurrence frequencies. Finally, the ranked numbers are transmitted to an entropy encoder. The proposed method improves the performance of existing entropy coding by transforming original gray-level values into rank based images using statistical co-occurrence frequencies of gray-level images. The simulation results, using gray-level images with 8-bits, show that the proposed method can reduce bit rate by up to 37.85% compared to existing conventional entropy coders.

• Journal of Distribution Science
• /
• v.15 no.5
• /
• pp.53-64
• /
• 2017

Purpose - This study focuses on use intention of mobile travel Apps by Chinese tourists visiting Korea based on UTAUT model, ISS model and ITM model. And the corresponding market promotion schemes are proposed for operators of mobile travel Apps by the research results. Research design, data, and methodology - After collecting 326 respondents in China with cross-sectional questionnaires, this study begins the empirical research with users of mobile travel Apps, and analyzes data with IBM SPSS 23.0 and IBM AMOS 23.0. Results - The results of this study include the following aspects: firstly, the System quality and Information quality are accepted for hypotheses of Satisfaction and Performance expectancy. Secondly, the Personal Propensity to Trust and Firm Reputation are accepted for Initial Trust hypothesis, and the hypotheses of Firm Reputation and Initial Trust are accepted for Use Intention. Thirdly, the Performance expectancy, Effort expectancy, Social influence are accepted for Use Intention hypothesis. Conclusions - With the increase of tourists visiting Korea, it can be predicted that the needs visiting Korea will be increased persistently for Chinese - this trend brings about the increase of the Chinese travel. First, information quality greatly influences satisfaction and performance expectancy. The research result shows that, the higher the mobile traveling App's information quality is, the higher the satisfaction and performance expectancy will be. Therefore, operators of mobile traveling App should have in-depth investigations towards users, to know the latter's real demand to the information quality and then provide corresponding services. Second, performance expectancy and effort expectancy greatly influence users' intention. Therefore, mobile traveling App operators should improve Apps' convenience and efficiency and, in doing so, find an effective method for market expansion. Third, social influence greatly affects users' intention. The result shows that mobile traveling App operators should pay attention to the influence of mass media and friends' recommendation on users, thereby it is necessary to improve advertisement activities. Fourth, initial trust also influences users' intention. The result shows that initial trust is a key element inducing users to generate use intention. Therefore, mobile traveling Apps operators should make efforts to catch elements that influence users' initial trust.

• Journal of the Korea Safety Management & Science
• /
• v.11 no.1
• /
• pp.103-113
• /
• 2009

Motorola which had developed and commercialized wireless telephone and mobile phone for the first time in the world made a decision on the separation of the mobile unit from the Motorola due to the continuous bad performance turn. In addition, the Korean company named Raincom which had developed and commercialized MP3 player have been disappeared since Raincom failed to compete with iPod of Apples. What is the main cause the companies that have been top of the world level weed out of IT product market? For less product life cycle and globalization of market, the company's competitiveness tends to be determined based on company's creative capability, rather than process implementation capability. Thus, companies spent a great amount of time and effort in the introduction of creativity management and R&D activities. However, as company's continuous investment on R&D and innovation activities did not lead to business performance, they starts to pay much attention on the effective R&D and innovation activities. Therefore, this paper tries to solve the limitation of the effectiveness of R&D activities by introducing the concept of 'creative quality'. The creative quality is executed by the creative process, which is designed for getting over the limitation of the traditional brainstorming method. This paper presents the concept so called 'creative quality management' and expects it to be an alternative to conquest the limitation of six sigma management.

• Smart Media Journal
• /
• v.11 no.4
• /
• pp.30-37
• /
• 2022

Disease threatens plant growth and recognizing the type of disease is essential to making a remedy. In recent years, deep learning has witnessed a significant improvement for this task, however, a large volume of labeled images is one of the requirements to get decent performance. But annotated images are difficult and expensive to obtain in the agricultural field. Therefore, designing an efficient and effective strategy is one of the challenges in this area with few labeled data. Transfer learning, assuming taking knowledge from a source domain to a target domain, is borrowed to address this issue and observed comparable results. However, current transfer learning strategies can be regarded as a supervised method as it hypothesizes that there are many labeled images in a source domain. In contrast, unsupervised transfer learning, using only images in a source domain, gives more convenience as collecting images is much easier than annotating. In this paper, we leverage unsupervised transfer learning to perform plant disease recognition, by which we achieve a better performance than supervised transfer learning in many cases. Besides, a vision transformer with a bigger model capacity than convolution is utilized to have a better-pretrained feature space. With the vision transformer-based unsupervised transfer learning, we achieve better results than current works in two datasets. Especially, we obtain 97.3% accuracy with only 30 training images for each class in the Plant Village dataset. We hope that our work can encourage the community to pay attention to vision transformer-based unsupervised transfer learning in the agricultural field when with few labeled images.

• Journal of the Korean Regional Science Association
• /
• v.38 no.1
• /
• pp.21-32
• /
• 2022

The purpose of this study is two things. First, it is to develop and demonstrate a sharong taxi platform. To this end, the implications for platform development were derived by analyzing consumers' perceptions of existing taxi services using IPA. As a result, abnormal business activities and safe services in the maintenance area were found to be safe rides and easy rides in the key improvement area. Safety such as usage fee level and driver information provision were derived in the areas subject to improvement, and friendly response and internal and external cleanliness were derived in the areas of excessive investment. The second purpose of this study is to estimate the value given to users for sharing taxi service platforms using the CVM. As a result of estimating the value of the demonstration service of the shared taxi platform developed through this study, the WTP was 3,621 won/per household/per year when expanding throughout Gimhae-si, and 2,515 won/per household/per year. Compared to the willingness to pay for empirical services, only 69.5% of the willingness to pay for the spread project in Gimhae-si. This is the result of a combination of service spread to an unspecified number of people and concerns about service quality due to spatial expansion. This suggests that it is necessary to build data through continuous demonstration and to carefully build a roadmap for spread by upgrading services based on this.

• Journal of the Korea society of information convergence
• /
• v.5 no.2
• /
• pp.49-58
• /
• 2012

Korea has entered into an aging society at an exceptionally high speed. Increasing aged workers give pressure on Korean companies with seniority based pay system with labor costs. The wage peak system has been introduced to Korean companies as an alternative to early retirements. The wage peak system currently adopted in Korea is mainly the aging-limit-security model. But in the future we should actively adopt the employment-extension model. To operate wage peak system successfully, I suggest introduction of the individual wage peak system, development of suitable jobs for the aged, selective application by job classification, incentive graded by performance evaluation, and working hour reduction with the decreased rate of wage.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.2
• /
• pp.610-630
• /
• 2020

A large number of people suffered from traffic accidents each year, so people pay more attention to traffic safety. However, the traditional methods use laser sensors to calculate the vehicle distance at a very high cost. In this paper, we propose a method based on deep learning to calculate the vehicle distance with a monocular camera. Our method is inexpensive and quite convenient to deploy on the mobile platforms. This paper makes two contributions. First, based on Light-Head RCNN, we propose a new vehicle detection framework called Light-Car Detection which can be used on the mobile platforms. Second, the planar homography of projective geometry is used to calculate the distance between the camera and the vehicles ahead. The results show that our detection system achieves 13FPS detection speed and 60.0% mAP on the Adreno 530 GPU of Samsung Galaxy S7, while only requires 7.1MB of storage space. Compared with the methods existed, the proposed method achieves a better performance.

• Korean System Dynamics Review
• /
• v.14 no.1
• /
• pp.5-29
• /
• 2013

This study tries to measure the direct and indirect effects of the Regional Strategic Industry (RSI) promotion projects in Chungbuk Province in Korea. In specific, it critically examines whether there exists policy consistency and connectivity between the hardware-oriented Stage I (2002~2007) and the software-centered Stage II (2008~2012) RSI promotion projects. Applying System Dynamics (SD) techniques, this study examines complex system characteristics of RSI promotion projects, all of which have been derived from the causal and stock-flow models and their simulated results. Major findings are as follows: Firstly, 'the continuous investment' is regarded as the most crucial policy leverage for the strategic industry promotion and regional economic growth. Secondly, without exceptions, the RSI promotion projects should switch their evaluation criteria to performance-oriented ones. Thirdly, in selecting their subprojects, the RSI promotion projects should pay due attention to evaluating technology value and marketability. Fourthly, it should put policy priority in strengthening cluster networking and interconnectivity among projects, inevitably supporting a selective number of virtuous network systems. Fifthly, auxiliary projects such as marketing, technology aid, and knowledge-based services should not be overlooked.

• Journal of Korea Multimedia Society
• /
• v.16 no.12
• /
• pp.1403-1412
• /
• 2013

In the maritime ships or airplanes, users should utilize the satellite channel in orer to use web service. However, the satellite channel is costly and does not give users satisfied response time. In the ship, the users may receive plenty of extra data when they obtain internet news. The extra data may be unnecessary image and advertise. Therefore, they should pay unnecessary data usage charges as well. In this paper, we suggest a proxy model that solves the problem of cost and speed. The proposed proxy reduces the of data through the satellite link by image and advertising blocking, caching, image re-requesting functions. It's performance was tested by a real satellite communication.

• Journal of Auto-vehicle Safety Association
• /
• v.9 no.1
• /
• pp.13-18
• /
• 2017

In recent years, NCAP regulations of many countries have induced automaker to improve the vehicle crashworthiness. But, the current NCAP regulations don't cover all types of traffic accidents. And rapid-increasing market share of compact cars and SUVs has brought for both consumer and automaker to pay more attention on crash compatibility. So, many countries have tried to develop the new crash test mode and update the present crash test mode. Especially, Euro NCAP has been developing a new impact protocol of the car-to-car frontal offset impact including the crash compatibility assessment. There are plans to introduce this new protocol in 2020, and it will be replaced the current Euro NCAP frontal offset impact. The test dummy in the front seats of this new test mode will be changed from 50% Hybrid-III male to 50% THOR male. This paper will address the vehicle responses, the occupant responses and the vehicle compatibility performance from a full vehicle crash test using the new car-to-car frontal offset test protocol of Euro NCAP.