• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• v.27 no.2
• /
• pp.146-159
• /
• 2023

This paper proposes a hybrid algorithm combining K-means clustering and watershed algorithms for flower segmentation and counting. We use the K-means clustering algorithm to obtain the main colors in a complex background according to the cluster centers and then take a color space transformation to extract pixel values for the hue, saturation, and value of flower color. Next, we apply the threshold segmentation technique to segment flowers precisely and obtain the binary image of flowers. Based on this, we take the Euclidean distance transformation to obtain the distance map and apply it to find the local maxima of the connected components. Afterward, the proposed algorithm adaptively determines a minimum distance between each peak and apply it to label connected components using the watershed segmentation with eight-connectivity. On a dataset of 30 images, the test results reveal that the proposed method is more efficient and precise for the counting of overlapped flowers ignoring the degree of overlap, number of overlap, and relatively irregular shape.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.10 no.2
• /
• pp.25-30
• /
• 2010

Wireless sensor networks is composed of many similar sensor nodes with limited resources. They are randomly scattered over a specific area and self-organize the network. For guarantee of network life time, load balancing and scalability in sensor networks, sensor networks needs the clustering algorithm which distribute the networks to a local cluster. In existing clustering algorithms, the cluster head selection method has two problems. One is additional communication cost for finding location and energy of nodes. Another is unequal clustering. To solve them, this paper proposes a novel cluster head selection algorithm revised previous clustering algorithm, LEACH. The simulation results show that the energy compared with the previous clustering method is reduced.

• ETRI Journal
• /
• v.35 no.2
• /
• pp.311-320
• /
• 2013

In most spectral clustering approaches, the Gaussian kernel-based similarity measure is used to construct the affinity matrix. However, such a similarity measure does not work well on a dataset with a nonlinear and elongated structure. In this paper, we present a new similarity measure to deal with the nonlinearity issue. The maximum flow between data points is computed as the new similarity, which can satisfy the requirement for similarity in the clustering method. Additionally, the new similarity carries the global and local relations between data. We apply it to spectral clustering and compare the proposed similarity measure with other state-of-the-art methods on both synthetic and real-world data. The experiment results show the superiority of the new similarity: 1) The max-flow-based similarity measure can significantly improve the performance of spectral clustering; 2) It is robust and not sensitive to the parameters.

• Journal of Korea Society of Industrial Information Systems
• /
• v.17 no.2
• /
• pp.57-64
• /
• 2012

In this paper, the multi-objective genetic algorithm is proposed for the document clustering which is important in the text mining field. The most important function in the document clustering algorithm is to group the similar documents in a corpus. So far, the k-means clustering and genetic algorithms are much in progress in this field. However, the k-means clustering depends too much on the initial centroid, the genetic algorithm has the disadvantage of coming off in the local optimal value easily according to the fitness function. In this paper, the multi-objective genetic algorithm is applied to the document clustering in order to complement these disadvantages while its accuracy is analyzed and compared to the existing algorithms. In our experimental results, the multi-objective genetic algorithm introduced in this paper shows the accuracy improvement which is superior to the k-means clustering(about 20 %) and the general genetic algorithm (about 17 %) for the document clustering.

• World Technopolis Review
• /
• v.1 no.2
• /
• pp.129-140
• /
• 2012

This paper presents, from a broad perspective, the manner in which various types of clusters and options for regional development have evolved in Korea over the past decade, with particular emphasis on who have taken initiative in establishing the clusters. Characterized by not only progress but also setbacks, two distinctive patterns have emerged: centralized de-concentration and regionalized concentration. Both the Korean government and numerous localities have continuously extended efforts to create different clusters, technology parks, special districts, etc. In many cases, local or regional governments have competed intensely for clusters to be located in their jurisdictions; in particular, concerted efforts to convince national governments to set up special districts have been witnessed. On the other hand, major localities have made their own efforts to generate large- and small-scale clustering projects. It remains to be seen how different outcomes or effectiveness these two approaches will make in the future. Following the review of relevant literature and practices, I examine the well-known national campaign and projects in the previous administration in Korea in the context of 'de-concentration' of economic values and resources. Thereafter, other cases initiated mostly by local governments are discussed; some of these clustering efforts and regional projects have fared well thus far, but some haven't. In the case of Daegu, the progress of some critical projects, such as the Daegu Technopolis and a Free Economic Zone, is elaborated.

• Smart Structures and Systems
• /
• v.5 no.1
• /
• pp.43-54
• /
• 2009

This study introduces a novel approach for locating damage in a structure using wireless sensor system with local level computational capability to alleviate data traffic load on the centralized computation. Smart wireless sensor systems, capable of iterative damage-searching, mimic an optimization process in a decentralized way. The proposed algorithm tries to detect damage in a structure by monitoring abnormal increases in strain measurements from a group of wireless sensors. Initially, this clustering technique provides a reasonably effective sensor placement within a structure. Sensor clustering also assigns a certain number of master sensors in each cluster so that they can constantly monitor the structural health of a structure. By adopting a voting system, a group of wireless sensors iteratively forages for a damage location as they can be activated as needed. Since all of the damage searching process occurs within a small group of wireless sensors, no global control or data traffic to a central system is required. Numerical simulation demonstrates that the newly developed searching algorithm implemented on wireless sensors successfully localizes stiffness damage in a plate through the local level reconfigurable function of smart sensors.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1998.10a
• /
• pp.504-510
• /
• 1998

In this paper, Fundamental rules governing group intelligence "obstacle avoidance" behavior of multiple autonomous mobile robots are represented by a small number of fuzzy rules. Complex lifelike behavior is considered as local interactions between simple individuals under small number of fundamental rules. The fuzzy rules for obstacle avoidance are generated from clustering the input-output data obtained from the obstacle avoidance algorithm. Simulation shows the fuzzy rules successfully realizes fundamental rules of the obstacle avoidance behavior.

• Journal of Korean Society on Water Environment
• /
• v.28 no.2
• /
• pp.202-212
• /
• 2012

Due to the deterioration of water supply network and the deficiency of raw water, the water utility of local governments have performed various projects to improve their revenue water ratio. However, it is very difficult to estimate the cost for maintaining the revenue water ratio at higher level after completing the project, because local governments have different conditions affecting the operating and maintenance cost of water supply network. The purpose of this study is to present a procedure to estimate the operating and maintenance cost required to maintain the target revenue water ratio of the water supply network. For this purpose, we estimated the cost used only for operation and maintenance of water supply network of 164 local governments with the aid of K-Mean Clustering Analysis and the data from 40 representative local governments. Then, the regression analysis was performed to find relationship between revenue water ratio and the operating and maintenance cost with two different data sets generated by two classification methods; the first method classifies the local governments by means of k-means clustering, and the other classifies the local governments according to the index standardized by the operating and maintenance cost per unit length of water mains per revenue water ratio. The results shows that the method based on the index standardized by the cost and revenue water ratio of each government produces more reliable results for finding regression equations between revenue water ratio and the operating and maintenance cost only for water supply network. The estimated regression equations for each group can be used to estimate the cost required to keep the target revenue water ratio of the local government.

• Journal of Korean Library and Information Science Society
• /
• v.40 no.2
• /
• pp.491-514
• /
• 2009

The supply of public libraries in Korea has been rapidly improving because of the recent increase for cultural demands and revitalization of the local culture. This paper aims to analyze the efficiency about 565 public libraries using Clustering DEA/AHP(CDA) model. This model is employed the efficiency analysis in order to incorporate project irreversibility and division due to the limit of resources spending. The results shows that the public libraries are divided into three groups which are large size libraries(Group 1), middle size libraries(Group 2) and small size libraries(Group 3). Their average efficiency was found as 0.89, 0.72 and 0.60 each.

• Journal of Communications and Networks
• /
• v.17 no.6
• /
• pp.568-581
• /
• 2015

Predicting locations of users with portable devices such as IP phones, smart-phones, iPads and iPods in public wireless local area networks (WLANs) plays a crucial role in location management and network resource allocation. Many techniques in machine learning and data mining, such as sequential pattern mining and clustering, have been widely used. However, these approaches have two deficiencies. First, because they are based on profiles of individual mobility behaviors, a sequential pattern technique may fail to predict new users or users with movement on novel paths. Second, using similar mobility behaviors in a cluster for predicting the movement of users may cause significant degradation in accuracy owing to indistinguishable regular movement and random movement. In this paper, we propose a novel fusion technique that utilizes mobility rules discovered from multiple similar users by combining clustering and sequential pattern mining. The proposed technique with two algorithms, named the clustering-based-sequential-pattern-mining (CSPM) and sequential-pattern-mining-based-clustering (SPMC), can deal with the lack of information in a personal profile and avoid some noise due to random movements by users. Experimental results show that our approach outperforms existing approaches in terms of efficiency and prediction accuracy.