• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.2
• /
• pp.631-653
• /
• 2020

Sign language is a natural, visually oriented and non-verbal communication channel between people that facilitates communication through facial/bodily expressions, postures and a set of gestures. It is basically used for communication with people who are deaf or hard of hearing. In order to understand such communication quickly and accurately, the design of a successful sign language translation system is considered in this paper. The proposed system includes object detection and classification stages. Firstly, Single Shot Multi Box Detection (SSD) architecture is utilized for hand detection, then a deep learning structure based on the Inception v3 plus Support Vector Machine (SVM) that combines feature extraction and classification stages is proposed to constructively translate the detected hand gestures. A sign language fingerspelling dataset is used for the design of the proposed model. The obtained results and comparative analysis demonstrate the efficiency of using the proposed hybrid structure in sign language translation.

• Journal of Environmental Health Sciences
• /
• v.40 no.3
• /
• pp.245-254
• /
• 2014

Objectives: This study was performed to investigate hand washing awareness of females and load of indicator bacteria on their hands. This study focused on the variation according to their age. Methods: A self-administered questionnaire survey and bacterial analysis of indicator bacteria were carried out for 100 Korean women in their age from 20 s to 60 s. Hand samples were collected through a modified glovejuice method. Results: In the survey, significant difference (p < 0.05) was found among the age groups in the use of hand washing agents. The levels of aerobic colony count (ACC) were the highest in both hands among the 20s (p < 0.05). The levels of Escherichia coli were higher in both hands in their 20s and 30s. No significant difference was found in the levels of Staphylococcus aureus and Salmonella spp. However, the positive rates of S. aureus (left hand, 37.5~47.1%; right hand, 58.5~62.5%) and Salmonella spp. (left hand, 25.0~52.9%; right hand, 37.5~64.7%) were higher in the hands of the 20s and 30s, and then showed decreasing trend according to increase of age. The effect of hand washing frequency on the ACC level of hands was significant (p < 0.001). Conclusions: These results indicate that there was no remarkable difference of hand hygiene awareness among female age groups. The detection of S. aureus and Salmonella spp. on the hands of some females in each age group revealed poor hand hygiene practices. The significant effect of hand washing frequency on the ACC level suggests that frequent hand washing is helpful to reduce hand contamination.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2016.07a
• /
• pp.297-298
• /
• 2016

본 논문에서는 손을 보다 효과적으로 인식하기 위해 동작 검출과 피부색 검출을 이용하여 인식하는 시스템을 제안한다. 단순히 피부색만을 이용하여 손을 인식하는 경우 피부색과 유사한 색상의 물체나 다른 신체 부위를 인식하는 문제점이 발생하게 된다. 이러한 문제점을 해결하기 위해 동작 검출을 이용하여 움직이는 물체만을 손이라고 가정하였다. 이렇게 가정을 하고 피부색 검출과 동작 검출을 이용하여 인식하는 경우 신체부위를 제외하고는 거의 검출되지 않는다. 그리고 인식된 영역마다 뼈대를 찾아 손을 검출한다. 조명이나 주변 환경에 최대한 영향을 적게 받기위해 시스템을 설계하였으며 단순 피부색 검출을 이용한 손 검출보다 좋은 성능을 발휘하며 손가락의 개수와 손 모양, 손 추적까지 응용할 수 있다.

• Structural Engineering and Mechanics
• /
• v.38 no.1
• /
• pp.81-97
• /
• 2011

In this research, the wavelet transform is used to analyze time response of a cracked beam carrying moving mass for damage detection. In this respect, a new damage detection method based on the combined use of continuous and discrete wavelet transforms is proposed. It is shown that this method is more capable in making damage signature evident than the traditional two approaches based on direct investigation of the wavelet coefficients of structural response. By the proposed method, it is concluded that strain data outperforms displacement data at the same point in revealing damage signature. In addition, influence of moving mass-induced terms such as gravitational, Coriolis, centrifuge forces, and pure inertia force along the deflection direction to damage detection is investigated on a sample case. From this analysis it is concluded that centrifuge force has the most influence on making both displacement and strain data damage-sensitive. The Coriolis effect is the second to improve the damage-sensitivity of data. However, its impact is considerably less than the former. The rest, on the other hand, are observed to be insufficient alone.

• Proceedings of the IEEK Conference
• /
• 2003.07a
• /
• pp.182-185
• /
• 2003

Error performance of two modulation schemes of millimeter-wave over fiber (MMoF) system i.e., optical double side band (ODSB) and optical single side band (OSSB) modulations is analyzed under Rician fading. Bit error rates (BER) of two detection techniques i.e., coherent and noncoherent detection are also compared in Rician fading. In aspect of error performance, ODSB modulation scheme has better BER than OSSB modulation scheme has under Rician fading. On the other hand, OSSB modulation scheme is advantageous in case of considering high bandwidth efficiency and small power degradation. Coherent detection technique is proper in Rician fading, because coherent detection provides more SNR gain whether fading is serious or not. Noncoherent detection can be applied when we need a simple receiver structure.

• Journal of Korea Multimedia Society
• /
• v.16 no.12
• /
• pp.1393-1402
• /
• 2013

Hand gesture is one of the most popular communication methods in everyday life. In human-computer interaction applications, hand gesture recognition provides a natural way of communication between humans and computers. There are mainly two methods of hand gesture recognition: glove-based method and vision-based method. In this paper, we propose a vision-based hand gesture recognition method using Kinect. By using the depth information is efficient and robust to achieve the hand detection process. The finger labeling makes the system achieve pose classification according to the finger name and the relationship between each fingers. It also make the classification more effective and accutate. Two kinds of gesture sets can be recognized by our system. According to the experiment, the average accuracy of American Sign Language(ASL) number gesture set is 94.33%, and that of general gestures set is 95.01%. Since our system runs in real-time and has a high recognition rate, we can embed it into various applications.

• The Journal of the Korea Contents Association
• /
• v.10 no.5
• /
• pp.20-28
• /
• 2010

In this paper, we propose a hand gesture interface for the manipulation of augmented objects in 3D space using a camera. Generally a marker is used for the detection of 3D movement in 2D images. However marker based system has obvious defects since markers are always to be included in the image or we need additional equipments for controling objects, which results in reduced immersion. To overcome this problem, we replace marker by planar hand shape by estimating the hand pose. Kalman filter is for robust tracking of the hand shape. The experimental result indicates the feasibility of the proposed algorithm for hand based AR interfaces.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.2
• /
• pp.855-870
• /
• 2020

Windows are key components of building facades. Detecting windows, crucial to 3D semantic reconstruction and scene parsing, is a challenging task in computer vision. Early methods try to solve window detection by using hand-crafted features and traditional classifiers. However, these methods are unable to handle the diversity of window instances in real scenes and suffer from heavy computational costs. Recently, convolutional neural networks based object detection algorithms attract much attention due to their good performances. Unfortunately, directly training them for challenging window detection cannot achieve satisfying results. In this paper, we propose an approach for window detection. It involves an improved Faster R-CNN architecture for window detection, featuring in a window region proposal network, an RoI feature fusion and a context enhancement module. Besides, a post optimization process is designed by the regular distribution of windows to refine detection results obtained by the improved deep architecture. Furthermore, we present a newly collected dataset which is the largest one for window detection in real street scenes to date. Experimental results on both existing datasets and the new dataset show that the proposed method has outstanding performance.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2009.05a
• /
• pp.321-325
• /
• 2009

The detection process of defective tags in most of Korean domestic RFID manufacturing companies is handled or treated by on-hand processing after the job of chip bonding, so it has been requesting to reduce the time and cost for manufacturing of RFID tags. Therefore, in this paper, we design and implement the system to perform the functionality of detection of defective tags after the process of chip bonding, and so provide the basis of a related software to establish the foundation of a automation system for the detection of defected RFID tags which is requested in the related Korean domestic industrial field. The developed system in this paper shows the enhancement of 700% in processing speed and 100% in detection rate of defective tags, comparing to the method of on-hand processing.

• IEIE Transactions on Smart Processing and Computing
• /
• v.1 no.2
• /
• pp.88-94
• /
• 2012

Many hand gesture recognition systems using advanced computer vision techniques to eliminate the need for a TV remote controller have been proposed. Nevertheless, some issues still remain, such as high computational complexity and insufficient information on the target object and background. Moreover, none of the proposed techniques consider how to enter the control mode of the system. This means that they may need a TV remote controller to enter the control mode. This paper proposes a hand mouse system using a pre-defined gesture with high background adaptability. By doing so, a remote controller to enter the control mode of the IPTV system can be eliminated.