• Journal of Digital Convergence
• /
• v.19 no.6
• /
• pp.251-258
• /
• 2021

In this paper, we proposed an encoder-decoder model utilizing residual learning to improve the accuracy of the U-Net-based semantic segmentation method. U-Net is a deep learning-based semantic segmentation method and is mainly used in applications such as autonomous vehicles and medical image analysis. The conventional U-Net occurs loss in feature compression process due to the shallow structure of the encoder. The loss of features causes a lack of context information necessary for classifying objects and has a problem of reducing segmentation accuracy. To improve this, The proposed method efficiently extracted context information through an encoder using residual learning, which is effective in preventing feature loss and gradient vanishing problems in the conventional U-Net. Furthermore, we reduced down-sampling operations in the encoder to reduce the loss of spatial information included in the feature maps. The proposed method showed an improved segmentation result of about 12% compared to the conventional U-Net in the Cityscapes dataset experiment.

• Journal of Korean Society of Forest Science
• /
• v.110 no.1
• /
• pp.72-81
• /
• 2021

We developed smartphone application and web application server to acquire and effectively manage tree measurement information. Smartphone applications can measure tree height, diameter at breast height (DBH), azimuth, altitude, slope, and positional coordinates using augmented reality (Google AR core) and motion sensors. The web application server effectively manages and stores measurement information. To evaluate the accuracy of information acquired using a smartphone, 90 Korean pine trees (Pinus koraiensis) were randomly selected from a natural mixed forest, with a total of 90 representative trees randomly collected from a natural mixed forest. Then, height and DBH were measured using a Haglof Vertex Laser Hypsometer and caliper. Comparisons of the results indicated significant results at the 95% level and a very high average correlation of 0.972 for both tree height and DBH. In terms of DBH, the average errors were 0.6745 cm and 1.0139 cm for artificial coniferous and natural mixed forests, respectively.

• KIPS Transactions on Computer and Communication Systems
• /
• v.11 no.9
• /
• pp.281-288
• /
• 2022

AGVs are often used in industrial applications to transport heavy materials around a large industrial building, such as factories or warehouses. In particular, in fulfillment centers their usefulness is maximized for automation. To increase productivity in warehouses such as fulfillment centers, sophisticated path planning of AGVs is required. We propose a scheme that can be applied to QMIX, a popular cooperative MARL algorithm. The performance was measured with three metrics in several fulfillment center layouts, and the results are presented through comparison with the performance of the existing QMIX. Additionally, we visualize the transport paths of trained AGVs for a visible analysis of the behavior patterns of the AGVs as heat maps.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.26 no.5D
• /
• pp.869-876
• /
• 2006

In the case of earthenware cultural heritages that are found in the form of fragments, the major task is quick and precise restoration. The existing method, which follows the rule of trial and error, is not only greatly time consuming but also lacked precision. If this job could be done by three dimensional scanning, matching up pieces could be done with remarkable efficiency. In this study, the original earthenware was modeled through three-dimensional pattern scanning and photogrammetry, and each of the fragments were scanned and modeled. In order to obtain images from the photogrammetry, we calibrated and used a Canon EOS 1DS real size camera. We analyzed the relationship among the sections of the formed model, efficiently compounded them, and analyzed the errors through residual and color error map. Also, we built a web-based three-dimensional simulation environment centering around the users, for the virtual museum.

• Journal of KIISE:Software and Applications
• /
• v.35 no.9
• /
• pp.572-587
• /
• 2008

The Self-organizing Feature Map(SOFM) that is one of unsupervised neural networks is a very powerful tool for data clustering and visualization in high-dimensional data sets. Although the SOFM has been applied in many engineering problems, it needs to cluster similar weights into one class on the trained SOFM as a post-processing, which is manually performed in many cases. The traditional clustering algorithms, such as t-means, on the trained SOFM however do not yield satisfactory results, especially when clusters have arbitrary shapes. This paper proposes automatic clustering on trained SOFM, which can deal with arbitrary cluster shapes and be globally optimized by graph cuts. When using the graph cuts, the graph must have two additional vertices, called terminals, and weights between the terminals and vertices of the graph are generally set based on data manually obtained by users. The Proposed method automatically sets the weights based on mode-seeking on a distance matrix. Experimental results demonstrated the effectiveness of the proposed method in texture segmentation. In the experimental results, the proposed method improved precision rates compared with previous traditional clustering algorithm, as the method can deal with arbitrary cluster shapes based on the graph-theoretic clustering.

• Journal of KIISE:Software and Applications
• /
• v.34 no.7
• /
• pp.646-654
• /
• 2007

In recent years, the use of text inserted into TV contents has grown to provide viewers with better visual understanding. In this paper, video text is defined as superimposed text region located of the bottom of video. Video text extraction is the first step for video information retrieval and video indexing. Most of video text detection and extraction methods in the previous work are based on text color, contrast between text and background, edge, character filter, and so on. However, the video text extraction has big problems due to low resolution of video and complex background. To solve these problems, we propose a method to extract text from videos using the Harris corner detector. The proposed algorithm consists of four steps: corer map generation using the Harris corner detector, extraction of text candidates considering density of comers, text region determination using labeling, and post-processing. The proposed algorithm is language independent and can be applied to texts with various colors. Text region update between frames is also exploited to reduce the processing time. Experiments are performed on diverse videos to confirm the efficiency of the proposed method.

• Journal of the Korean Institute of Landscape Architecture
• /
• v.45 no.2
• /
• pp.23-39
• /
• 2017

This study began from the question, "is there a way to efficiently apply industrial demand in the university curriculum?" Research focused on how to actively accept and respond to the era of the NCS (National Competency Standards). In order to apply NCS to individual departments of the university, industrial personnel must positively participate to form a practical-level curriculum by the NCS, which can be linked to the work and qualifications. A valid procedure for developing a curriculum based on the NCS of this study is as follows: First, the university must select a specific classification of NCS considering the relevant industry outlook, the speciality of professors in the university, the relationship with regional industries and the prospects for future employment, and the need for industrial manpower. Second, departments must establish a type of human resource that compromises goals for the university education and the missions of the chosen NCS. In this process, a unique competency unit of the university that can support the basic or applied subjects should be added to the task model. Third, the task model based on the NCS should be completed through the verification of each competency unit considering the acceptance or rejection in the curriculum. Fourth, subjects in response to each competency units within the task model should be developed while considering time and credits according to university regulations. After this, a clear subject description of how to operate and evaluate the contents of the curriculum should be created. Fifth, a roadmap for determining the period of operating subjects for each semester or year should be built. This roadmap will become a basis for the competency achievement frame to decide upon the adoption of a Process Evaluation Qualification System. In order for the NCS to be successfully established within the university, a consensus on the necessity of the NCS should be preceded by professors, students and staff members. Unlike a traditional curriculum by professors, the student-oriented NCS curriculum is needed sufficient understanding and empathy for the many sacrifices and commitment of the members of the university.

• Particle and aerosol research
• /
• v.6 no.3
• /
• pp.107-121
• /
• 2010

The number of nanotechnology based consumer products are growing rapidly. Thus, the customer likely to be exposed to such products continues to increase as the applications expand. This article describes the international and Korea's policies on the EHS(Environment, Safety and Health) issues of nanotechnologies. The strategic plan and coordination of OECD and ISO were summarized. This article also examines several new findings of Korean researchers as well as current and future challenges in the aerosol application study of EHS issues on the nanotechnologies.

• Journal of Navigation and Port Research
• /
• v.40 no.4
• /
• pp.207-212
• /
• 2016

The box-counting(BC) method is one of the most commonly used methods for fractal dimension calculation of binary images in the fields of Engineering, Science, Medical Science, Geology, etc due to its simplicity and reliability. It deals with only square images with each size equal to the power of 2 to prevent it from discarding unused pixels for images of arbitrary size. In this paper, we presents a more efficient BC method based on the original one, which is applicable to images of arbitrary size. The proposed approach allows the number of the counting boxes to be real to improve the estimation accuracy. The mean absolute error performance is computed on two deterministic fractal images whose theoretical dimensions are well known to compare with those of the existing BC method and triangular BC method. The experimental results show that the proposed method can outperform the two methods and assess the complexity of coastline images of Korea and Chodo island taken from the Google map.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.13 no.11
• /
• pp.2251-2259
• /
• 2009

We propose an algorithm and system which generates 3D stereo image by composition of 2D image from 4 multiple clusters which 1 cluster was composed of 4 multiple cameras based on network. Proposed Schemes have a network-based client-server architecture for load balancing of system caused to process a large amounts of data with real-time as well as multiple cluster environments. In addition, we make use of JPEG compression and RAM disk method for better performance. Our scheme first converts input images from 4 channel, 16 cameras to binary image. And then we generate 3D stereo images after applying edge detection algorithm such as Sobel algorithm and Prewiit algorithm used to get disparities from images of 16 multiple cameras. With respect of performance results, the proposed scheme takes about 0.05 sec. to transfer image from client to server as well as 0.84 to generate 3D stereo images after composing 2D images from 16 multiple cameras. We finally confirm that our scheme is efficient to generate 3D stereo images in multiple view and multiple clusters environments with real-time.