• Smart Structures and Systems
• /
• v.29 no.1
• /
• pp.153-166
• /
• 2022

Identifying fine cracks in steel bridge facilities is a challenging task of structural health monitoring (SHM). This study proposed an end-to-end crack image segmentation framework based on a one-step Convolutional Neural Network (CNN) for pixel-level object recognition with high accuracy. To particularly address the challenges arising from small object detection in complex background, efforts were made in loss function selection aiming at sample imbalance and module modification in order to improve the generalization ability on complicated images. Specifically, loss functions were compared among alternatives including the Binary Cross Entropy (BCE), Focal, Tversky and Dice loss, with the last three specialized for biased sample distribution. Structural modifications with dilated convolution, Spatial Pyramid Pooling (SPP) and Feature Pyramid Network (FPN) were also performed to form a new backbone termed CrackDet. Models of various loss functions and feature extraction modules were trained on crack images and tested on full-scale images collected on steel box girders. The CNN model incorporated the classic U-Net as its backbone, and Dice loss as its loss function achieved the highest mean Intersection-over-Union (mIoU) of 0.7571 on full-scale pictures. In contrast, the best performance on cropped crack images was achieved by integrating CrackDet with Dice loss at a mIoU of 0.7670.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.49 no.11
• /
• pp.64-73
• /
• 2012

In this paper, we propose a method to search for identical videos. The proposed method is spatio-temporal group feature fingerprinting. Frame of video is extracted from fixed rate method and is partitioned into vertical group and horizontal group. Descriptor is made of each group feature that is extracted from binary fingerprinting. Next, use descriptor of original video to build a two type of fingerprinting database and matching with query video. To efficient and effective video copy detection, method have high robustness, independence, matching speed. In proposed method, group feature have high robustness and independence in variable modification of video. Building a original fingerprinting database is able to fast matching with query video. The proposed method shows performance improvement in variable modifications in comparison to the existing methods. Especially, very singular performance in speed improvement is great advantage of this paper.

• The Korean Journal of Air & Space Law and Policy
• /
• v.7
• /
• pp.151-175
• /
• 1995

In recent years, questions have arisen under several forms with respect to the need for adapting present legal order established under the Chicago Convention and relevant customary rules into newly developed environment surrounding the international air transport industry. Major feature of such trends included in opinions for modification of the present legal order might be defined as more liberalistic approach to this industry. In this respect, many scholars and lawyers in this field agree with a view that a theoretical tie between an aircraft/air liners and a register - State lies in political and strategical concern of the State so that each aircraft/air liners has been attributed a single nationality. In the context of such concern, each aircraft/air liners has been related with each register-State in the form of "genuine connection". However, present and near future development of air transport industry and its world - wide market requires some modification of such single nationality regime. Taking into account such circumstances, States as creator of present legal order are in the process of establishing new legal order where air liners with multi - nationality are capable of satisfying to such needs. As adopting a series of liberalization package for air transport industry in european continent, European Union adopts a concept of "community air carrier", by which an air space of each member State is open to each other, especially through the grant of cabotage right. A serious concern may arise in such grant because the Article 7 of the Chicago Convention prohibits such grant on an exclusive basis. While many theoretical opinions have been put forward concerning the interpretation of that article, a case of European Union shall be a good test of the range of its application. It is anticipated that future development around this issue shaH furnish us a major feature of the liberalization of international air transportation and an adaptation process of present legal order.

• Journal of Computational Design and Engineering
• /
• v.3 no.2
• /
• pp.161-177
• /
• 2016

In history based parametric CAD modeling systems, persistent identification of the topological entities after design modification is mandatory to keep the design intent by recording model creation history and modification history. Persistent identification of geometric and topological entities is necessary in the product design phase as well as in the re-evaluation stage. For the identification, entities should be named first according to the methodology which will be applicable for all the entities unconditionally. After successive feature operations on a part body, topology based persistent identification mechanism generates ambiguity problem that usually stems from topology splitting and topology merging. Solving the ambiguity problem needs a complex method which is a combination of topology and geometry. Topology is used to assign the basic name to the entities. And geometry is used for the ambiguity solving between the entities. In the macro parametrics approach of iCAD lab of KAIST a topology based persistent identification mechanism is applied which will solve the ambiguity problem arising from topology splitting and also in case of topology merging. Here, a method is proposed where no geometry comparison is necessary for topology merging. The present research is focused on the enhancement of the persistent identification schema for the support of ambiguity problem especially of topology splitting problem and topology merging problem. It also focused on basic naming of pattern features.

• Transactions on Electrical and Electronic Materials
• /
• v.10 no.6
• /
• pp.193-195
• /
• 2009

This paper aims to develop a low gate source voltage ($V_{gs}$) N-LDMOS element that is fully operational at a CMOS Logic Gate voltage (3.3 or 5 V) realized using the 0.35 μm BCDMOS process. The basic structure of the N-LDMOS element presented here has a Low $V_{gs}$ LDMOS structure to which the thickness of a logic gate oxide is applied. Additional modification has been carried out in order to obtain features of an improved breakdown voltage and a specific on resistance ($R_{sp}$). A N-LDMOS element can be developed with improved features of breakdown voltage and specific on resistance, which is an important criterion for power elements by means of using a proper structure and appropriate process modification. In this paper, the structure has been made to withstand the excessive electrical field on the drain side by applying the double gate oxide structure to the channel area, to improve the specific on resistance in addition to providing a sufficient breakdown voltage margin. It is shown that the resulting modified N-LDMOS structure with the feature of the specific on resistance is improved by 31%, and so it is expected that optimized power efficiencies and the size-effectiveness can be obtained.

• The International Journal of Costume Culture
• /
• v.3 no.2
• /
• pp.161-174
• /
• 2000

Today's advanced transport and mass media enables people to access easily and frequently to traditional costumes from different countries in the world. Consequently, fashion designers who are fascinated by the variety and beauty of traditional costumes, tend to adapt them to modern fashion design. The continuous study of traditional costumes at such time contains a meaning as a source of inspiration for the creation of a new fashion design. Therefore, among various costumes of the world, this study will focus on the costume of Palestine, that is situated at the crossroads of Europe, Asia and Afria and has exchanged various historical cultures with other cultural regions. It, especially, expects to provide a unique idea in the material modification field of today's fashion design by examining mainly the embroidery ornament which is the most distinctive feature of Palestinian costume. The conclusion was revealed that Palestinian women's costume largely consists of embroidered silk and cotton garments with a V-slitted square panel (gabeh) attached to the chest part. And a variety of luxurious embroidery was employed to emphasize different parts of the dress. These embroideries differ from area to area. Especially the embroidery from Bethlehem was well-known in many places and influenced those of other regions.

• Earthquakes and Structures
• /
• v.4 no.6
• /
• pp.607-627
• /
• 2013

The application of shape memory alloys (SMAs) to the seismic response reduction of civil engineering structures has attracted growing interest due to their self-centering feature and excellent fatigue performance. The loading rate dependence of SMAs raises a concern in the seismic analysis of SMA-based devices. However, the implementation of micromechanics-based strain-rate-dependent constitutive models in structural analysis software is rather complicated and computationally demanding. This paper investigates the feasibility of replacing complex rate-dependent models with rate-independent constitutive models for superelastic SMA elements in seismic time-history analysis. Three uniaxial constitutive models for superelastic SMAs, including one rate-dependent thermomechanical model and two rate-independent phenomenological models, are considered in this comparative study. The pros and cons of the three nonlinear constitutive models are also discussed. A parametric study of single-degree-of-freedom systems with different initial periods and strength reduction factors is conducted to examine the effect of the three constitutive models on seismic simulations. Additionally, nonlinear time-history analyses of a three-story prototype steel frame building with special SMA-based damping braces are performed. Two suites of seismic records that correspond to frequent and design basis earthquakes are used as base excitations in the seismic analyses of steel-braced frames. The results of this study show that the rate-independent constitutive models, with their parameters properly tuned to dynamic test data, are able to predict the seismic responses of structures with SMA-based seismic response modification devices.

• Journal of KIISE:Software and Applications
• /
• v.34 no.2
• /
• pp.140-149
• /
• 2007

This paper presents a statistical model for Korean dependency-based parsing. Although Korean is one of free word order languages, it has the feature of which some word order is preferred to local contexts. Earlier works proposed parsing models using modification lengths due to this property. Our model uses headible path contexts for modification length probabilities. Using a headible path of a dependent it is effective for long distance relation because the large surface context for a dependent are abbreviated as its headible path. By combined with lexical bigram dependency, our probabilistic model achieves 86.9% accuracy in eojoel analysis for KAIST corpus, more improvement especially for long distance dependencies.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.32 no.1C
• /
• pp.78-85
• /
• 2007

With the advancement of the digital broadcasting and popularity of the Internet, recently, many studies are making on the digital watermarking for the copyright protection of digital data. This paper proposes the minimum modification method of motion vector to minimize the degradation of video quality, hiding subtitles of many language and information of OST(original sound track), character profiles, etc. as well as the copyright protection. Our proposed algorithm extracts feature vector by comparing motion vector data with watermark data, and minimize the modification of motion vectors by deciding the inversion of bit. Thus the degradation of video quality is minimized comparing to conventional algorithms. This algorithm also can check data integrity, and retrieve embedded hidden data simply and blindly. And our proposed scheme can be useful for conventional MPEG-1, -2 standards without any increment of bit rate in the compressed video domain. The experimental result shows that the proposed scheme obtains better video quality than other previous algorithms by about $0.5{\sim}1.5dB$.

• KIPS Transactions on Computer and Communication Systems
• /
• v.3 no.5
• /
• pp.163-172
• /
• 2014

The Scale Invariant Feature Transform (SIFT) has been widely used in a lot of applications for image feature matching. Such a transform allows us to strong matching ability, stability in rotation, and scaling with the variety of different scales. Recently, it has been made one of the most successful algorithms in the research areas of copy-move forgery detections. Though this transform is capable of identifying copy-move forgery, it does not widely address the possibility that counter-forensics operations may be designed and used to hide the evidence of image tampering. In this paper, we propose a targeted counter-forensics method for impeding SIFT-based copy-move forgery detection by applying a semantically admissible distortion in the processing tool. The proposed method allows the attacker to delude a similarity matching process and conceal the traces left by a modification of SIFT keypoints, while maintaining a high fidelity between the processed images and original ones under the semantic constraints. The efficiency of the proposed method is supported by several experiments on the test images with various parameter settings.