• Journal of the Korea Society of Computer and Information
• /
• v.20 no.6
• /
• pp.87-97
• /
• 2015

An Aerial video summarization is not only the key to effective browsing video within a limited time, but also an embedded cue to efficiently congregative situation awareness acquired by unmanned aerial vehicle. Different with previous works, we utilize sensor operation mode of unmanned aerial vehicle, which is global, local, and focused surveillance mode in order for accurately summarizing the aerial video considering flight and surveillance/reconnaissance environments. In focused mode, we propose the moving-react tracking method which utilizes the partitioning motion vector and spatiotemporal saliency map to detect and track the interest moving object continuously. In our simulation result, the key frames are correctly detected for aerial video summarization according to the sensor operation mode of aerial vehicle and finally, we verify the efficiency of video summarization using the proposed mothed.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.43 no.6 s.312
• /
• pp.28-35
• /
• 2006

Statistical models of shape variability based on active shape models (ASMs) have been successfully utilized to perform segmentation and recognition tasks in two-dimensional (2D) images. Three-dimensional (3D) model-based approaches are more promising than 2D approaches since they can bring in more realistic shape constraints for recognizing and delineating the object boundary. For 3D model-based approaches, however, building the 3D shape model from a training set of segmented instances of an object is a major challenge and currently it remains an open problem in building the 3D shape model, one essential step is to generate a point distribution model (PDM). Corresponding landmarks must be selected in all1 training shapes for generating PDM, and manual determination of landmark correspondences is very time-consuming, tedious, and error-prone. In this paper, we propose a novel automatic method for generating 3D statistical shape models. Given a set of training 3D shapes, we generate a 3D model by 1) building the mean shape fro]n the distance transform of the training shapes, 2) utilizing a tetrahedron method for automatically selecting landmarks on the mean shape, and 3) subsequently propagating these landmarks to each training shape via a distance labeling method. In this paper, we investigate the accuracy and compactness of the 3D model for the human liver built from 50 segmented individual CT data sets. The proposed method is very general without such assumptions and can be applied to other data sets.

• The KIPS Transactions:PartD
• /
• v.10D no.1
• /
• pp.1-12
• /
• 2003

Recently, there have been many indexing schemes for multimedia data such as image, video data. But recent database applications, for example data mining and multimedia database, are required to support multi-user environment. In order for indexing schemes to be useful in multi-user environment, a concurrency control algorithm is required to handle it. So we propose a concurrency control algorithm that can be applied to CBF (cell-based filtering method), which uses the signature of the cell for alleviating the dimensional curse problem. In addition, we extend the SHORE storage system of Wisconsin university in order to handle high-dimensional data. This extended SHORE storage system provides conventional storage manager functions, guarantees the integrity of high-dimensional data and is flexible to the large scale of feature vectors for preventing the usage of large main memory. Finally, we implement the web-based image retrieval system by using the extended SHORE storage system. The key feature of this system is platform-independent access to the high-dimensional data as well as functionality of efficient content-based queries. Lastly. We evaluate an average response time of point query, range query and k-nearest query in terms of the number of threads.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.9 no.2
• /
• pp.231-242
• /
• 2014

It is throwing the spotlight on the cultivation crops about high quality crops and productivity improvement per unit area because of rapid climate change caused by global warming. Therefore, we propose a water culture management of circulation nutrient method control system applies to solar tracking method not using traditional method of deep flow technique and artificial light source. We design it in the form of the circulation nutrient method in waterway of a certain amount of nutrient solution and water flowed into the way of circular. In addition, we design a multistage structure in pyramid shape which be possible continuous photosynthesis action to crops of water culture bottom part. Also, solar tracking method is designed five sensor method of center hole sensor method for tracking shadow of solar light not using traditional two hole, four hole sensor method. Finally, through the water culture device applies to solar light tracking method was not introduced in existing study yet, we can reduce growth speed of crops which be possible continuous photosynthesis action to crops. Moreover, We can expect high productivity of per unit area which be possible all crops can be offered growth environment of same type by using form of pyramid shape of multistage structure without top or bottom part.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.29 no.6
• /
• pp.598-608
• /
• 2023

To comprehensively grasp the dynamic changes in the coastal terrain and coastal erosion, it is imperative to incorporate temporal and spatial continuity through frequent and continuous monitoring. Recently, there has been a proliferation of research in coastal monitoring using remote sensing, accompanied by advancements in image monitoring and analysis technologies. Remote sensing, typically involves collection of images from aircraft or satellites from a distance, and offers distinct advantages in swiftly and accurately analyzing coastal terrain changes, leading to an escalating trend in its utilization. Remote satellite image-based coastal line detection involves defining measurable coastal lines from satellite images and extracting coastal lines by applying coastal line detection technology. Drawing from the various data sources surveyed in existing literature, this study has comprehensively analyzed encompassing the definition of coastal lines based on satellite images, current status of remote satellite imagery, existing research trends, and evolving landscape of technology for satellite image-based coastal line detection. Based on the results, research directions, on latest trends, practical techniques for ideal coastal line extraction, and enhanced integration with advanced digital monitoring were proposed. To effectively capture the changing trends and erosion levels across the entire Korean Peninsula in future, it is vital to move beyond localized monitoring and establish an active monitoring framework using digital monitoring, such as broad-scale satellite imagery. In light of these results, it is anticipated that the coastal line detection field will expedite the progression of ongoing research practices and analytical technologies.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.30 no.2
• /
• pp.147-156
• /
• 2024

The sustainable management and enhancement of marine resources are becoming increasingly important issues worldwide. This study was conducted in response to these challenges, focusing on the development and performance comparison of fish detection and classification models as part of a deep learning-based technique for assessing the effectiveness of marine resource enhancement projects initiated by the Korea Fisheries Resources Agency. The aim was to select the optimal model by training various sizes of YOLOv8-Seg models on a fish image dataset and comparing each performance metric. The dataset used for model construction consisted of 36,749 images and label files of 12 different species of fish, with data diversity enhanced through the application of augmentation techniques during training. When training and validating five different YOLOv8-Seg models under identical conditions, the medium-sized YOLOv8m-Seg model showed high learning efficiency and excellent detection and classification performance, with the shortest training time of 13 h and 12 min, an of 0.933, and an inference speed of 9.6 ms. Considering the balance between each performance metric, this was deemed the most efficient model for meeting real-time processing requirements. The use of such real-time fish detection and classification models could enable effective surveys of marine resource enhancement projects, suggesting the need for ongoing performance improvements and further research.

• Korean Journal of Construction Engineering and Management
• /
• v.15 no.4
• /
• pp.11-19
• /
• 2014

Construction VE activities has occupies an important position as a key tool of the project in terms of budget and performance improvements. Reflecting this situation, VE application scope is being expanded than before. Thus, construction projects are required the continued research and development for effective use of construction VE. VE target choice method is the central part in the construction VE work. But, this has a problem with low utilization in practical terms. Accordingly, this study was developed the new methodology(PVTCM) for VE target selection of the construction VE work. Its contents are as follows: First, to analyze practical problem for VE target choice method through VE practical data. Second, to extract the high potential VE object based on the cost model and ideas similarity by excellent VE cases analysis. Third, to implement standardized VE object item by using affinity diagram for each construction type items. This research were conducted to experts interviews for comparison analysis with the cost model as one of a representative target selection method. As a result, proposed method were implemented significantly higher performance in terms of practical effectiveness than conventional method.

• Journal of Korea Multimedia Society
• /
• v.13 no.11
• /
• pp.1715-1727
• /
• 2010

Although a hand haptic interaction which provides direct and natural sensation is the most natural way of interacting with VR environment, the hand haptic interaction has still limitations with respect to the complexity of articulated hand and related hardware capabilities. Particularly, virtual assembly simulation which refers to the verification process of digital mockup in product development lifecycle is one of the most challenging topics in virtual reality applications. However, hand haptic interaction is considered as a big obstacle, because difficulty initial grasping and non-dextrous manipulation remain as unsolved problems. In this paper, we propose that common hand haptic interactions involves two separate stages with different aspects. We present the hand haptic interaction method enables us to stably grasp a virtual object at initial grasping and delicately manipulate an object at task operating by one's intention. Therefore, proposed method provides the robustness using grasping quality and dextrous manipulation using physically simulation. We conducted experiments to evaluate the effectiveness of our proposed method under different display environments -monoscopic and stereoscopic. From 2-way ANOVA test, we show that the proposed method satisfies two aspects of hand haptic interaction. Finally, we demonstrated an actual application of various assembly simulation for relatively complex models.

• Journal of Intelligence and Information Systems
• /
• v.21 no.1
• /
• pp.15-28
• /
• 2015

As personal devices and pervasive technologies for interacting with networked objects continue to proliferate, there is an unprecedented world of scattered pieces of contextualized information available. However, the explosive growth and variety of information ironically lead users and service providers to make poor decision. In this situation, recommender systems may be a valuable alternative for dealing with these information overload. But they failed to utilize various types of contextual information. In this study, we suggest a methodology for context-aware recommender systems based on the concept of contextual boundary. First, as we suggest contextual boundary-based profiling which reflects contextual data with proper interpretation and structure, we attempt to solve complexity problem in context-aware recommender systems. Second, in neighbor formation with contextual information, our methodology can be expected to solve sparsity and cold-start problem in traditional recommender systems. Finally, we suggest a methodology about context support score-based recommendation generation. Consequently, our methodology can be first step for expanding application of researches on recommender systems. Moreover, as we suggest a flexible model with consideration of new technological development, it will show high performance regardless of their domains. Therefore, we expect that marketers or service providers can easily adopt according to their technical support.

• Tunnel and Underground Space
• /
• v.30 no.5
• /
• pp.462-472
• /
• 2020

Rock classification is fundamental discipline of exploring geological and geotechnical features in a site, which, however, may not be easy works because of high diversity of rock shape and color according to its origin, geological history and so on. With the great success of convolutional neural networks (CNN) in many different image-based classification tasks, there has been increasing interest in taking advantage of CNN to classify geological material. In this study, a feasibility of the deep CNN is investigated for automatically and accurately identifying rock types, focusing on the condition of various shapes and colors even in the same rock type. It can be further developed to a mobile application for assisting geologist in classifying rocks in fieldwork. The structure of CNN model used in this study is based on a deep residual neural network (ResNet), which is an ultra-deep CNN using in object detection and classification. The proposed CNN was trained on 10 typical rock types with an overall accuracy of 84% on the test set. The result demonstrates that the proposed approach is not only able to classify rock type using images, but also represents an improvement as taking highly diverse rock image dataset as input.