• Smart Structures and Systems
• /
• v.30 no.6
• /
• pp.613-626
• /
• 2022

Guaranteeing the quality and integrity of structural health monitoring (SHM) data is very important for an effective assessment of structural condition. However, sensory system may malfunction due to sensor fault or harsh operational environment, resulting in multiple types of data anomaly existing in the measured data. Efficiently and automatically identifying anomalies from the vast amounts of measured data is significant for assessing the structural conditions and early warning for structural failure in SHM. The major challenges of current automated data anomaly detection methods are the imbalance of dataset categories. In terms of the feature of actual anomalous data, this paper proposes a data anomaly detection method based on data-level and deep learning technique for SHM of civil engineering structures. The proposed method consists of a data balancing phase to prepare a comprehensive training dataset based on data-level technique, and an anomaly detection phase based on a sophisticatedly designed network. The advanced densely connected convolutional network (DenseNet) and Transformer encoder are embedded in the specific network to facilitate extraction of both detail and global features of response data, and to establish the mapping between the highest level of abstractive features and data anomaly class. Numerical studies on a steel frame model are conducted to evaluate the performance and noise immunity of using the proposed network for data anomaly detection. The applicability of the proposed method for data anomaly classification is validated with the measured data of a practical supertall structure. The proposed method presents a remarkable performance on data anomaly detection, which reaches a 95.7% overall accuracy with practical engineering structural monitoring data, which demonstrates the effectiveness of data balancing and the robust classification capability of the proposed network.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.6
• /
• pp.43-52
• /
• 2022

We can sense somebody's feeling fatigue, which means that fatigue can be detected through sensing human biometric signals. Numerous researches for assessing fatigue are mostly focused on diagnosing the edge of disease-level fatigue. In this study, we adapt quantitative analysis approaches for estimating qualitative data, and propose video analysis models for measuring fatigue state. Proposed three deep-learning based classification models selectively include stages of video analysis: object detection, feature extraction and time-series frame analysis algorithms to evaluate each stage's effect toward dividing the state of fatigue. Using frontal face videos collected from various fatigue situations, our CNN model shows 0.67 accuracy, which means that we empirically show the video analysis models can meaningfully detect fatigue state. Also we suggest the way of model adaptation when training and validating video data for classifying fatigue.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.26 no.12A
• /
• pp.2109-2119
• /
• 2001

In this paper, a fast information extraction algorithm for object-based MPEG-4 application from MPEG-2 bit-steam is proposed. For object-based MPEG-4 conversion, we need to extract such information as object-image, shape-image, macro-block motion vector, and header information from MPEG-2 bit-stream. If we use the extracted information, fast conversion for object-based MPEG-4 is possible. The proposed object extraction algorithm has two important steps, namely the motion vectors extraction from MPEG-2 bit-stream and the watershed algorithm. The algorithm extracts objects using user\`s assistance in the intra frame and tracks then in the following inter frames. If we have an unsatisfactory result for a fast moving object, the user can intervene to correct the segmentation. The proposed algorithm consist of two steps, which are intra frame object extracts processing and inter frame tracking processing. Object extracting process is the step in which user extracts a semantic object directly by using the block classification and watersheds. Object tacking process is the step of the following the object in the subsequent frames. It is based on the boundary fitting method using motion vector, object-mask, and modified watersheds. Experimental results show that the proposed method can achieve a fast conversion from the MPEG-2 bit-stream to the object-based MPEG-4 input.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.41 no.3
• /
• pp.91-102
• /
• 2004

In this paper, a fast information extraction algorithm for object-based MPEG-4 application from MPEG-1,2 is proposed. For object-based MPEG-4 conversion, we need to extract such information as object-image, shape-image, macro-block motion vector, and header information from MPEG-1,2 bit-stream. If we use the extracted information, fast conversion for object-based MPEG-4 is possible. The proposed object extraction algerian has two important steps, namely the motion vector extraction from MPEG-1,2 bit-stream and the watershed algerian The algorithm extracts objects using user's assistance in the intra frame and tracks then in the following inter frames. If we have an unsatisfactory result for a fast moving object the user can intervene to connect the segmentation. The proposed algorithm consist of two steps, which are intra frame object extracting processing and inter frame tracking processing. Object extracting process is the step in which user extracts a semantic object directly by using the block classification and watersheds. Object tracking process is the step of the following the object in the subsequent frames. It is based on the boundary fitting method using motion vector, object-mask and modified watersheds. Experimental results show that the proposed method can achieve a fast conversion from the MPEG-1,2 bit-stream to the object-based MPEG-4 input.

• Proceedings of the Acoustical Society of Korea Conference
• /
• autumn
• /
• pp.81-84
• /
• 1999

본 논문에서는 중규모급 단어인식기의 실시간 구현을 위한 무감독 단어집단화 알고리듬을 제안한다. 무감독 단어집단화는 인식대상 어휘 수가 많은 대용량 음성인식 시스템에서 대상 어휘 수를 줄여주는 역할을 하는 전처리기의 성격을 갖는다. 무감독 집단화를 위해 각 단어의 유$\cdot$무성음 고유의 특성을 잘 반영할 수 있는 특징 파라미터 5개를 사용하여 패턴 인식과 회귀분석에서 널리 사용되고 있는 분류$\cdot$회귀트리(Classification And Regression Tree)에 적용시키는 방법으로 접근하였고, 각 단어의 frame 수를 일정하게 n개로 분할(segment)하여 1개의 tree를 생성시키는 방법과 각 segment에 해당하는 tree를 생성시켜 segment들 사이의 교집합 성분으로 단어들을 집단화 하였다 실험결과 탐색 대상단어 22개에서 평균2.21개로 줄어 전체 대상 단어의 $10\%$만을 탐색하여 인식할 수 있는 방법을 제시할 수 있었다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.10 no.8
• /
• pp.3806-3825
• /
• 2016

This paper brings out a neoteric frame of reference for visual semantic based 3d video search and retrieval applications. Newfangled 3D retrieval application spotlight on shape analysis like object matching, classification and retrieval not only sticking up entirely with video retrieval. In this ambit, we delve into 3D-CBVR (Content Based Video Retrieval) concept for the first time. For this purpose we intent to hitch on BOVW and Mapreduce in 3D framework. Here, we tried to coalesce shape, color and texture for feature extraction. For this purpose, we have used combination of geometric & topological features for shape and 3D co-occurrence matrix for color and texture. After thriving extraction of local descriptors, TB-PCT (Threshold Based- Predictive Clustering Tree) algorithm is used to generate visual codebook. Further, matching is performed using soft weighting scheme with L₂ distance function. As a final step, retrieved results are ranked according to the Index value and produce results .In order to handle prodigious amount of data and Efficacious retrieval, we have incorporated HDFS in our Intellection. Using 3D video dataset, we fiture the performance of our proposed system which can pan out that the proposed work gives meticulous result and also reduce the time intricacy.

• Journal of Electrical Engineering and Technology
• /
• v.11 no.6
• /
• pp.1857-1862
• /
• 2016

Human activity recognition using depth information is an emerging and challenging technology in computer vision due to its considerable attention by many practical applications such as smart home/office system, personal health care and 3D video games. This paper presents a novel framework of 3D human body detection, tracking and recognition from depth video sequences using spatiotemporal features and modified HMM. To detect human silhouette, raw depth data is examined to extract human silhouette by considering spatial continuity and constraints of human motion information. While, frame differentiation is used to track human movements. Features extraction mechanism consists of spatial depth shape features and temporal joints features are used to improve classification performance. Both of these features are fused together to recognize different activities using the modified hidden Markov model (M-HMM). The proposed approach is evaluated on two challenging depth video datasets. Moreover, our system has significant abilities to handle subject's body parts rotation and body parts missing which provide major contributions in human activity recognition.

• The KIPS Transactions:PartB
• /
• v.16B no.4
• /
• pp.271-280
• /
• 2009

In this paper, we propose a smoke detection method using region growing method in outdoor video sequences. Our proposed method is composed of three steps; the initial change area detection step, the boundary finding and expanding step, and the smoke classification step. In the first step, we use a background subtraction to detect changed areas in the current input frame against the background image. In difference images of the background subtraction, we calculate a binary image using a threshold value and apply morphology operations to the binary image to remove noises. In the second step, we find boundaries of the changed areas using labeling algorithm and expand the boundaries to their neighbors using the region growing algorithm. In the final step, ellipses of the boundaries are estimated using moments. We classify whether the boundary is smoke by using the temporal information.

• Journal of Korea Game Society
• /
• v.17 no.5
• /
• pp.133-142
• /
• 2017

Recently, screen climbing contents such as sports climbing learning program and screen climbing games. Especially, there are many researches on screen climbing games. In this paper, we propose the skeleton correction method based on the body area of a climber to improve the posture recognition accuracy. The correction method consists of the modified skeletal frame normalization with abnormal skeleton joint filtering, the classification of body area into joint parts, and the final skeleton joint correction. The skeletal information obtained by the proposed method can be used to compare the climber's posture and the ideal climbing posture.

• Journal of the Korean housing association
• /
• v.21 no.5
• /
• pp.23-33
• /
• 2010

In the late twentieth century, the multistory housing has become the most preferred housing type. Moreover, in multistory housing design, not only interior design but also exterior design, uniqueness and characteristic have been added as significant factors in design to break simplicity and uniformity. The purpose of this study is to understand the tendency of the elevation design in Europe and to suggest the idea for elevation design for domestic. Also, each elements of elevation were analyzed through classification and schematization which categorized by its own characteristic. Design elements can be classified into three categories. They are corresponding with inner spaces, emphasizing visual expression of exterior material and texture, and presenting aesthetic factors of building shape. Those are named as the Spatial Externalization, the Visual Expression, and the Expression of Building Shape. In this frame nineteen elevations of sixteen different cases were analysed. As a result of case studies, it is clear that elevation designs were not standardized and each cases has its own characteristics. Also it showed how the each design elements can be coordinated as a total design and how they expressed identities of each housing. This study could contribute to motivate diversifying the design of multistory housing.