• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.32B no.12
• /
• pp.1591-1602
• /
• 1995

A hybrid approach which includes edge- and region-based methods is considered. The modified non-linear Laplacian(MNL) filter is used for feature extraction. The matching algorithm has three steps which are edge, signed region, and residual region matching. At first, the edge points are matched using the sign and direction of edges. Then, the disparity is propagated from edge to inside region. A variable window is used to consider the local method which give accurate matched points and area-based method which can obtain full-resolution disparity map. In addition, a new relaxation algorithm for considering matching possibility derived from normalized error and regional continuity constraint is proposed to reduce the mismatched points. By the result of simulation for various images, this algorithm is insensitive to noise and gives full- resolution disparity map.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2019.06a
• /
• pp.32-34
• /
• 2019

최근 딥러닝 기반의 다양한 물체 검출 알고리즘이 제안되어 높은 성능을 보이고 있다. 본 논문은 이러한 딥러닝 기반 물체 검출의 성능을 향상시키기 위해 입력영상에서 추출된 특징 지도를 강화하는 비지역적 특징 융합과, 이를 이용한 물체 검출 기법을 제안한다. 제안 기법은 입력영상에서 CNN 을 통해 추출한 특징 지도를 비지역적 특징 강화 블록을 통해 강화한다. 해당 블록 내에서 입력된 특징 지도는 먼저 여러 리셉티브 필드를 갖는 특징 지도로 분기된다. 그리고 분기된 특징 지도들은 비지역적 특징 융합 모듈에 의해 융합되어 강화된다. 이러한 과정을 통해 강화된 특징 지도는 비지역적 문맥 정보가 강화된 특성을 가지며, 해당 특징 지도를 이용하여 최종적으로 물체 검출을 수행한다. Pascal VOC 공인 데이터세트를 통한 실험 결과, 제안 기법은 기존 비교 기법 대비 향상된 검출 성능을 보인다.

• Journal of Distribution Science
• /
• v.17 no.9
• /
• pp.25-33
• /
• 2019

Purpose - The purpose of the study was to seek to find out what factors having differences between paid membership customers (Costco membership) and general customers in retail industry. Since Costco operates differently from other conventional retailers, which is expected to have a substantial impact on consumers' preference of retail stores. Research design, data, and methodology - The survey conducted covered 1,000 adults in their 30s~50s living in Goyang and Gwangmyeong where Costco runs stores to determine the effects of Costco's local market-entry from consumer perspectives. 500 respondents were surveyed in each region and those working in the retail sector were excluded to ensure the objectivity of the answers. Results - Costco members in Goyang considered the price, bulk purchasing, and membership benefits as important criteria when choosing their retail store. On the other hand, as for Gwangmyeong, the non-member group's prominent characteristic was that they considered accessibility including travel distance and location and in-store amenities including food court services as important criteria for decision-making. Conclusion - Unique business model of Costco shows a statistically significant difference in terms of consumer awareness. the feature of Costco served as an critical criteria for consumers in their purchasing decision. Moreover, Bulk packaging purchases at Costco results in a strong supplementary relationship with neighborhood supermarkets.

• Journal of the Korean Geographical Society
• /
• v.29 no.2
• /
• pp.117-136
• /
• 1994

Contemporary global space economy is so dynamic that any one specific structural force can not explain the whole dynamic processes or trajectories of spatial industrial development. The major purpose of this paper is extending the traditional notion of industrial districts to functioning and development of new industrial districts with relation to the development of high technology industries. Several dynamic forces, which are dominated in new industrial districts in the modern space economy, are incorporated in the formation and dynamic aspects of new industrial districts. Even though key forces governing Marshallian industrial district are localization of small firms, division of labor between firms, constructive cooperation, and industrial atmosphere, Marshall points out a possibility of growing importance of large firms and non-local networks in the districts with changes of external environments. Some of Italian industrial districts can be regarded as Marshallian industrial districts in broader context, but the role of local authorities or institutions and local embeddedness seem to be more important in the Italian industrial districts. More critical implication form the review of Marshallian industrial districts and Italian industrial districts is that the industrial districts are not a static concept but a dynamic one: small firm based industrial districts can be regarded as only a specific feature evolved over time. Dynamic aspects of new industrial districts are resulting from coexistence of contrasting forces governing the functioning and formation of the districts in contemporary global space economy. The contrasting forces governing new industrial districts are coexistence of flexible and mass production systems, local and global networks, local and non-local embeddedness, and small and large firms. Because of these coexistence of contrasting forces, there are various types of new industrial districts. Nine types of industrial districts are identified based on local/non-local networks and intensity of networks in both suppliers and customers linkages. The different types of new industrial districts are described by differences in production systems, embeddedness, governance, cooperation and competition, and institutional factors. Out of nine types of industrial districts, four types - Marshallian; suppliers hub and spoke; customers hub and spoke; and satellite - are regarded as distinctive new industrial districts and four additional types - advanced hub and spoke types (suppliers and customers) and mature satellites (suppliers and customers) - can be evolved from the distinctive types and may be regarded as hybrid types. The last one - pioneering high technology industrial district - can be developed from the advanced hub and spoke types and this type is a most advanced modern industrial district in the era of globalization and high technology. The dynamic aspects of the districts are related with the coexistence of the contrasting forces in the contemporary global space economy. However, the development trajectory is not a natural one and not all the industrial districts can develop to the other hybrid types. Traditionally, localization of industries was developed by historical chances. In the process of high technology industrial development in contemporary global space economy, however, policy and strategies are critical for the formation and evolution of new industrial districts. It needs formation of supportive tissues of institutions for evolution of dyamic pattern of high technology related new industrial districts. Some of the original distinctive types of new industrial districts can not follow the path or trajectory suggested in this paper and may be declined without advancing, if there is no formation of supportive social structure or policy. Provision of information infrastructure and diffusion of an entrepreneurship through the positive supports of local government, public institutions, universities, trade associations and industry associations are important for the evolution of the dynamic new industrial districts. Reduction of sunk costs through the supports for training and retraining of skilled labor, the formation of flexible labor markets, and the establishment of cheap and available telecommunication networks is also regarded as a significant strategies for dynamic progress of new industrial districts in the era of high technology industrial development. In addition, development of intensive international networks in production, technology and information is important policy issue for formation and evolution of the new industrial districts which are related with high technology industrial development.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.23 no.6
• /
• pp.682-689
• /
• 2019

In part-based image representation, the partial shapes of an object are represented as basis vectors, and an image is decomposed as a linear combination of basis vectors where the coefficients of those basis vectors represent the partial (or local) feature of an object. In this paper, a face recognition for occluded faces is proposed in which face images are represented using non-negative matrix factorization(NMF), one of part-based representation techniques, and recognized using an artificial neural network technique. Standard NMF, projected gradient NMF and orthogonal NMF were used in part-based representation of face images, and their performances were compared. Learning vector quantizer were used in the recognizer where Euclidean distance was used as the distance measure. Experimental results show that proposed recognition is more robust than the conventional face recognition for the occluded faces.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.8 no.5
• /
• pp.1676-1689
• /
• 2014

At present, the Content-Based Image Retrieval (CBIR) system has become a hot research topic in the computer vision field. In the CBIR system, the accurate extractions of low-level features can reduce the gaps between high-level semantics and improve retrieval precision. This paper puts forward a new retrieval method aiming at the problems of high computational complexities and low precision of global feature extraction algorithms. The establishment of the new retrieval method is on the basis of the SIFT and Harris (APISH) algorithm, and the salient region of interest points (SRIP) algorithm to satisfy users' interests in the specific targets of images. In the first place, by using the IPDSH and SRIP algorithms, we tested stable interest points and found salient regions. The interest points in the salient region were named as salient interest points. Secondary, we extracted the pseudo-Zernike moments of the salient interest points' neighborhood as the feature vectors. Finally, we calculated the similarities between query and database images. Finally, We conducted this experiment based on the Caltech-101 database. By studying the experiment, the results have shown that this new retrieval method can decrease the interference of unstable interest points in the regions of non-interests and improve the ratios of accuracy and recall.

• Korean Journal of Remote Sensing
• /
• v.31 no.4
• /
• pp.293-302
• /
• 2015

Water body extraction is significant for flood disaster monitoring using satellite imagery. Conventional methods have focused on finding an index, which highlights water body and suppresses non-water body such as vegetation or soil area. The Normalized Difference Water Index (NDWI) is typically used to extract water body from satellite images. The drawback of NDWI, however, is that some man-made objects in built-up areas have NDWI values similar to water body. The objective of this paper is to propose a new method that could extract correctly water body with built-up areas in before and after images of flood. We first create a two-element feature vector consisting of NDWI and a Near InfRared band (NIR) and then select a training site on water body area. After computing the mean vector and the covariance matrix of the training site, we classify each pixel into water body based on Mahalanobis distance. We also register before and after images of flood using outlier removal and triangulation-based local transformation. We finally create a change map by combining the before-flooding water body and after-flooding water body. The experimental results show that the overall accuracy and Kappa coefficient of the proposed method were 97.25% and 94.14%, respectively, while those of the NDWI method were 89.5% and 69.6%, respectively.

• Journal of Korean Society of Occupational and Environmental Hygiene
• /
• v.24 no.1
• /
• pp.79-90
• /
• 2014

Objectives: The aim of this study was to suggest preliminary data for the establishment of communication methodology of asbestos risk, fit for the features of each audiences, by grasping the features of risk communication by each element for each group survey. Methods: For this study, a questionnaire survey has been conducted from May to August 2012 and responses of 617 people including 214 school asbestos managers, 95 asbestos business managers, and 308 general public have been analyzed. Results: The feature by element of risk communication shows that to give information through non-governmental organizations with reliability such as colleges, research institutes, asbestos-related associations, etc among the entire investigated groups, is most effective. Lastly, for stakeholders related to asbestos, the public feedback for governmental asbestos management policy shows that it was considered that there is lack of reality due to comprehension deficit for situation, lack of a system of asbestos general management in the country and lack of policy connectivity among the branches of the government, and between the central government and the local government. However, the general public selected lack of various information disclosure, education, publicity for asbestos and lack of communication with citizens as the biggest problems.

• Journal of Korea Multimedia Society
• /
• v.10 no.2
• /
• pp.179-186
• /
• 2007

This paper proposes a novel method for face detection method using Zernike moments. To detect the faces in an image, local regions in multiscale sliding windows are classified into face and non-face by a neural network, and input features of the neural network consist of Zernike moments. Feature dimension is reduced as the reconstruction capability of orthogonal moment. In addition, because the magnitude of Zernike moment is invariant to rotation, a tilted human face can be detected. Even so the detection rate of the proposed method about head on face is less than experiments using intensity features, the result of our method about rotated faces is more robust. If the additional compensation and features are utilized, the proposed scheme may be best suited for the later stage of classification.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.12
• /
• pp.3868-3888
• /
• 2022

A widely used social networking service like Twitter has the ability to disseminate information to large groups of people even during a pandemic. At the same time, it is a convenient medium to share irrelevant and unverified information online and poses a potential threat to society. In this research, conventional machine learning algorithms are analyzed to classify the data as either non-rumor data or rumor data. Machine learning techniques have limited tuning capability and make decisions based on their learning. To tackle this problem the authors propose a deep learning-based Rumor Detection Neural Network model to predict the rumor tweet in real-world events. This model comprises three layers, AttCNN layer is used to extract local and position invariant features from the data, AttBi-LSTM layer to extract important semantic or contextual information and HPOOL to combine the down sampling patches of the input feature maps from the average and maximum pooling layers. A dataset from Kaggle and ground dataset #gaja are used to train the proposed Rumor Detection Neural Network to determine the veracity of the rumor. The experimental results of the RDNN Classifier demonstrate an accuracy of 93.24% and 95.41% in identifying rumor tweets in real-time events.