• Journal of Crop Science and Biotechnology
• /
• v.11 no.3
• /
• pp.163-170
• /
• 2008

The detection, characterization and use of quantitative traits loci, QTL, have significant potential to improve the efficiency of selective breeding of species. Therefore, a population with 59 advanced backcross lines($BC_2F_5$), derived from a cross between IR64 and Tarome molaei, were studied in Tonekabon Rice Research Station of Iran in order to map QTLs for panicle length, number of grain per panicle, and panicle grain sterility in rice. The parental screening wtih 235 SSR markers in agarose and polyacrylamide gels revealed 114 markers with clear polymorphic bands. To search for QTLs associated with panicle length, number of grain per panicle, and panicle grain sterility, we constructed a genetic linkage map using 114 microsatellite markers. Positive and negative transgressive segregations were observed in $BC_2F_5$ lines for all traits. Using multiple interval mapping(MIM), a total of 20 putative QTLs were detected, of which eight were for panicle length, three for number of grains, and nine for panicle grain sterility. The maximum number of QTLs were mapped on chromosomes 1 and 2 with eight QTLs. These QTL markers could possible be utilized for marker-assisted selection.

• Journal of Broadcast Engineering
• /
• v.25 no.7
• /
• pp.1081-1094
• /
• 2020

With the recent development of deep learning, most computer vision-related tasks are being solved with deep learning-based network technologies such as CNN and RNN. Computer vision tasks such as object detection or object segmentation use intermediate features extracted from the same backbone such as Resnet or FPN for training and inference for object detection and segmentation. In this paper, an experiment was conducted to find out the compression efficiency and the effect of encoding on task inference performance when the features extracted in the intermediate stage of CNN are encoded. The feature map that combines the features of 256 channels into one image and the original image were encoded in HEVC to compare and analyze the inference performance for object detection and segmentation. Since the intermediate feature map encodes the five levels of feature maps (P2 to P6), the image size and resolution are increased compared to the original image. However, when the degree of compression is weakened, the use of feature maps yields similar or better inference results to the inference performance of the original image.

• Journal of Korean Institute of Architectural Sustainable Environment and Building Systems
• /
• v.12 no.6
• /
• pp.557-566
• /
• 2018

The ZEB certification system has come into effect since 2018 according to the ZEB road map. From 2020, the public buildings with the total floor area smaller than $3000m^2$ are the target of the certification, and it will be extended to the buildings with up to $5000m^2$ floor area in 2025. However, current mandatory regulations for public office buildings seem already to meet the ZEB certification system planned for 2025. In this work, two buildings belong to $3,000{\sim}5,000m^2$ in total floor area were selected to analyze the possibility of meeting the ZEB certification only by following current obligation regulations. Results showed that the test buildings satisfied the minimum building energy efficiency and energy self-sufficiency rates for the ZEB certification when the mandatory insulation and installation ratio of renewable energy are applied. This can be useful for revising the road map or extending the target buildings of the ZEB certification.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.8
• /
• pp.1996-2015
• /
• 2023

High-definition (HD) maps can provide precise road information that enables an autonomous driving system to effectively navigate a vehicle. Recent research has focused on leveraging semantic segmentation to achieve automatic annotation of HD maps. However, the existing methods suffer from low recognition accuracy in automatic driving scenarios, leading to inefficient annotation processes. In this paper, we propose a novel semantic segmentation method for automatic HD map annotation. Our approach introduces a new encoder, known as the convolutional transformer hybrid encoder, to enhance the model's feature extraction capabilities. Additionally, we propose a multi-level fusion module that enables the model to aggregate different levels of detail and semantic information. Furthermore, we present a novel decoupled boundary joint decoder to improve the model's ability to handle the boundary between categories. To evaluate our method, we conducted experiments using the Bird's Eye View point cloud images dataset and Cityscapes dataset. Comparative analysis against stateof-the-art methods demonstrates that our model achieves the highest performance. Specifically, our model achieves an mIoU of 56.26%, surpassing the results of SegFormer with an mIoU of 1.47%. This innovative promises to significantly enhance the efficiency of HD map automatic annotation.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.18 no.8
• /
• pp.2298-2315
• /
• 2024

This paper introduces a method aimed at diagnosing the presence or absence of lesions by detecting anomalies in Chest X-ray images. The proposed approach is based on the PatchCore anomaly detection method, which extracts a feature vector containing location information of an image patch from normal image data and calculates the anomaly distance from the normal vector. However, applying PatchCore directly to medical image processing presents challenges due to the possibility of diseases occurring only in specific organs and the presence of image noise unrelated to lesions. In this study, we present an image alignment method that utilizes affine transformation parameter prediction to standardize already captured X-ray images into a specific composition. Additionally, we introduce a region-specific abnormality detection method that requires affine-transformed chest X-ray images. Furthermore, we propose a method to enhance application efficiency and performance through feature map hard masking. The experimental results demonstrate that our proposed approach achieved a maximum AUROC (Area Under the Receiver Operating Characteristic) of 0.774. Compared to a previous study conducted on the same dataset, our method shows a 6.9% higher performance and improved accuracy.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.13 no.4
• /
• pp.101-110
• /
• 2010

The increased occurrence of flooding due to typhoons and local rainfall has necessitated damage prevention through the systematic construction of damage history and quantitative analysis of flood prediction data. In this study, we constructed a disaster information map for practical use by combining digital images and continuous cadastral maps of damaged areas using a geographic information system to provide basic data and attribute information. In addition, we predicted the areas at risk of flash floods by calculating the flood capacity of the study area for different rainfall frequencies through flood inundation simulation, which was used to obtain comprehensive disaster information. Further, we calculated the extent of the flooded area and the damage rate for different rainfall frequencies using cadastral information. Flood inundation simulation in the case of heavy rainfall was found to help improve the ability to react to a flood and enhance the efficiency of rescue work by supporting decision-making for disaster management.

• Journal of KIISE:Software and Applications
• /
• v.32 no.4
• /
• pp.237-246
• /
• 2005

Product Line Engineering(PLE) is one of the technologies that develop applications economically reusing core assets. PLE consists of Framework Engineering(FE) and Application Engineering. Framework Engineering is to develop core assets that have common functionality shared by a set of family members. Application Engineering is to develop a specific application by instantiating the core assets. The PLE process increases reusability and efficiency because a specific application is developed by using core assets with less time and effort. Since definition of PLE artifacts and relationship between artifacts are not clear. developers have several troubles to make artifacts based on PLE process, are difficult to maintain consistency between artifacts, and do not use PLE process more practically. In this paper, we define meta-models of artifacts that are produced in PLE activities of PLE process and describe the traceability relationship between artifacts by using traceability map and guidelines that can apply traceability relationship. Finally, we define the way how trace links and guidelines of traceability map are applied.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.33 no.5
• /
• pp.427-436
• /
• 2015

This study aims to select an optimal site for the development of small-scaled artificial ground water recharge system with the purpose of 50ton/day pumping in dry season. First of all, the topography shape satisfying the numerous factors of a hydraulic model experiment is defined and an appropriate pumping efficiency is calculated through the model experiment of injection and pumping scenario. In next step, GIS(Geographic Information System) database are constructed by processing several geospatial data to explore the optimal site. In detail, watershed images are generated from DEM(Digital Elevation Model) with 5m cell size, which is set for the minimum area of the optimal site selection. Slope maps are made from DEM to determine the optimal hydraulic gradient to procure the proper aquifer undercurrent period. Finally, the suitable site for artificial recharge system is selected using an integration of overall data, such as an alluvial map, DEM, orthoimages, slope map, and watershed images.

• Journal of the Korean Society for information Management
• /
• v.22 no.3 s.57
• /
• pp.37-53
• /
• 2005

The terminology management is absolutely necessary for maintaining the efficiency of thesaurus. This is because the creating, differentiating, disappearing, and other processes of the descriptor become accomplished dynamically, making effective management of thesaurus a very difficult task. Therefore, a device is required for accomplishing methods to construct and maintain the thesaurus. This study proposes the methods to construct the thesaurus management using the basic elements of a topic map which are topic, occurrence, and association. Second, the study proposes the methods to represent the basic and specific instances using the systematic mapping algorithm and merging algorithm. Also, using a hub document as a standard, this study gives the methods to expand and subsitute the descriptors using the topic type. The new method applying fixed concept for double layer management on terms is developed, too. The purpose of this method is to fix the conceptual term which represents independent concept of time and space, and to select the descriptor freely by external information circumstance.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.6 no.6
• /
• pp.909-914
• /
• 2011

In these days, according to the enhancement of the mobile communication technologies, location based services using the location and movements are vitalized. In this paper, we develop a private vehicle tracking system for garbage truck operation management using GPS and CDMA communication and Open map interface. The terminal equipment attached in vehicle receives the GPS signals and detect the position data including time, longitude, latitude, and altitude. And the terminal sends these data to the server PC through CDMA cellular network in fixed period. The server saves these data into database to process in map server program with which we can view real-time trace of moving vehicle. We apply our system to public garbage truck managing operation and we can increase operation efficiency by examine real-time working and moving path and by present reasonable operation rout.