• Journal of Broadcast Engineering
• /
• v.17 no.5
• /
• pp.734-741
• /
• 2012

A new lossless intra coding method based on residual transform is applied to the next generation video coding standard HEVC (High Efficiency Video Coding). HEVC includes a multi-directional spatial prediction method to reduce spatial redundancy by using neighboring samples as a prediction for the samples in a block of data to be encoded. In the new lossless intra coding method, the spatial prediction is performed as samplewise DPCM (Difference Pulse Code Modulation) but is implemented as block-based manner by using residual transform and secondary residual transform on the HEVC standard. Experimental results show that the new lossless intra coding method reduces the bit rate by approximately 6.45% in comparison with the lossless intra coding method previously included in the HEVC standard.

• Journal of the Korean Society for Precision Engineering
• /
• v.28 no.5
• /
• pp.537-542
• /
• 2011

The thermo-mechanical interaction between brake block and wheel tread during braking has been found to cause thermal crack on the wheel tread. Due to thermal expansion of the rim material, the thermal cracks will protrude from the wheel tread and be more exposed to wear during the wheel/block contact than the rest of the tread surface. The wheel rim is in residual compression stress when is new. After service running, the region in the tread has reversed to tension. This condition can lead to the formation and growth of thermal cracks in the rim which can ultimately lead to premature failure of wheel. In the present paper, the thermal cracks of railway wheel, one of severe damages on the wheel tread, were evaluated to understand the safety of railway wheel in running condition. The residual stresses for damaged wheel which are applied to tread brake are investigated. Mainly X-ray diffusion method is used. Under the condition of concurrent loading of continuous rolling contact with rails and cyclic frictional heat from brake blocks, the reduction of residual stress is found to correlate well with the thermal crack initiation.

• Journal of Ocean Engineering and Technology
• /
• v.22 no.6
• /
• pp.95-99
• /
• 2008

The block assembly of ship consists of a certain type of heat processes such as cutting, bending welding residual stress relaxation and fairing. The residual deformation due to welding is inevitable at each assembly stage. The geometric inaccuracy caused by the welding deformation tends to preclude the introduction of automation and mechanization and needs the additional man-hours for the adjusting work at the following assembly stage. To overcome this problem, a distortion control method should be applied. For this purpose, it is necessary to develop an accurate prediction method which can explicitly account for the influence of various factors on the welding deformation. The validity of the prediction method must be also clarified through experiments. This paper proposes a simplified analysis method to predict the welding deformation of panel block structure. For this purpose, a simple prediction model for fillet welding deformations has been derived based on numerical and experimental results through the regression analysis. On the basis of these results, the simplified analysis method has been applied to some examples to show its validity.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.11a
• /
• pp.559-561
• /
• 2022

인공지능 시대에 들어서면서 개인 맞춤형 환경을 제공하기 위하여 사람의 감정을 인식하고 교감하는 기술이 많이 발전되고 있다. 사람의 감정을 인식하는 방법으로는 얼굴, 음성, 신체 동작, 생체 신호 등이 있지만 이 중 가장 직관적이면서도 쉽게 접할 수 있는 것은 표정이다. 따라서, 본 논문에서는 정확도 높은 얼굴 감정 식별을 위해서 Convolution Block Attention Module(CBAM)의 각 Gate와 Residual Block, Skip Connection을 이용한 Identity- CBAM Module을 제안한다. CBAM의 각 Gate와 Residual Block을 이용하여 각각의 표정에 대한 핵심 특징 정보들을 강조하여 Context 한 모델로 변화시켜주는 효과를 가지게 하였으며 Skip-Connection을 이용하여 기울기 소실 및 폭발에 강인하게 해주는 모듈을 제안한다. AI-HUB의 한국인 감정 인식을 위한 복합 영상 데이터 세트를 이용하여 총 6개의 클래스로 구분하였으며, F1-Score, Accuracy 기준으로 Identity-CBAM 모듈을 적용하였을 때 Vanilla ResNet50, ResNet101 대비 F1-Score 0.4~2.7%, Accuracy 0.18~2.03%의 성능 향상을 달성하였다. 또한, Guided Backpropagation과 Guided GradCam을 통해 시각화하였을 때 중요 특징점들을 더 세밀하게 표현하는 것을 확인하였다. 결과적으로 이미지 내 표정 분류 Task에서 Vanilla ResNet50, ResNet101을 사용하는 것보다 Identity-CBAM Module을 함께 사용하는 것이 더 적합함을 입증하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.9
• /
• pp.2882-2903
• /
• 2022

Image super-resolution (SR) processing is of great value in the fields of digital image processing, intelligent security, film and television production and so on. This paper proposed a densely connected deep learning network based on cascade architecture, which can be used to solve the problem of super-resolution in the field of image quality enhancement. We proposed a more efficient residual scaling dense block (RSDB) and the multi-channel cascade architecture to realize more efficient feature reuse. Also we proposed a hybrid loss function based on L₁ error and L_∞ error to achieve better L_∞ error performance. The experimental results show that the overall performance of the network is effectively improved on cascade architecture and residual scaling. Compared with the residual dense net (RDN), the PSNR / SSIM of the new method is improved by 2.24% / 1.44% respectively, and the L_∞ performance is improved by 3.64%. It shows that the cascade connection and residual scaling method can effectively realize feature reuse, improving the residual convergence speed and learning efficiency of our network. The L_∞ performance is improved by 11.09% with only a minimal loses of 1.14% / 0.60% on PSNR / SSIM performance after adopting the new loss function. That is to say, the L_∞ performance can be improved greatly on the new loss function with a minor loss of PSNR / SSIM performance, which is of great value in L_∞ error sensitive tasks.

• Maxillofacial Plastic and Reconstructive Surgery
• /
• v.29 no.5
• /
• pp.444-450
• /
• 2007

In general, labiolingual or buccolingual widths of residual alveolar bone are insufficient in edentulous area, because of alveolar resorption. Horizontal augmentation is bone graft procedure with a view to reinforcing horizontally insufficient bone quantity for installation of implants. The standard method is taking appropriate amount of block bone from intraoral or extraoral autogenous bone, and solid fixation with screws or mini-plate on labial or buccal side of residual alveolar bone. The purpose of this study is to discuss clinical usefulness of horizontal augmentation with autogenous block bone by observation and analysis of course of 41 implants installed to 12 patients by horizontal augmentation in Seoul National University Bundang Hospital from July, 2002 to December, 2005. The mean age of patients is 52.7, from 19 to 70, and the number of men and women is each 2 and 10. Block bone was taken from symphysis, body, ramus of mandible or iliac bone. And 6 types of implants were installed simultaneously or not, the diameters of implants are from 3.3 to 5.5mm, the lengths are from 8 to 15mm. The operator added artificial bone grafting material and optionally covered with membrane. The mean periods of observation after operation and final prosthetics were 28.6 and 17.0 months. As a result, 40 among 41 implants survived, the survival rate was 97.6%. Average 0.9mm crestal resorption was observed at final point of time by periapical view of each patients. Major complication related to the procedure was numbness in 7 patients.

• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.22 no.8
• /
• pp.659-664
• /
• 2009

This paper describes lightning impulse operating duty performance and temperature dependence of series gap type for transmission line arresters. The ageing parameters of lightning arresters are impulse current, moisture ingress, temperature ageing and so on. Especially it is important to estimate the change of electrical characteristics by lightning impulse current. In the discharge withstand test, total energy applied to the ZnO arrester each time is 4/10 ${\mu}s$, 30 kA. and in the operating duty test, the arrester has passed the test if thermal stability is achieved, if the residual voltage measured before and after the test is not changed by more than 5 %, and after the test reveals no evidence of puncture, flashover or cracking of the ZnO block. As a results, the residual voltage was in the range of 17.2${\sim}$20.3 kV and ZnO block bear up against at 2 shot of series impulse current of 30 kA. Also it was so excellent that the mechanical destruction does not occur at the 2 groups of 5 impulses current of 2/20 ${\mu}s$ 10 kA. According to the tests, it is thought that the ZnO arrester shows good stability with impulse current test. and it was found that the ambient temperature is increased resistive leakage current was increased in the range 47.3${\sim}$167.4 ${\mu}A$.

• Proceedings of the KSRS Conference
• /
• 2003.11a
• /
• pp.1168-1170
• /
• 2003

The GMS geometric correction method with highspeed and high accuracy is needed. In this paper, a selection method of residual errors for the GMS geometric correction using GCPs (ground control points) is described. Namely, it is a technique for limiting the number of residual error acquisition using GCPs in each block to reduce the processing time. As the result, since the processing time was about 7.0 minutes on conventional geometric correction and about 5.6 minutes on the proposed method, it was shown that the processing time of about 1.4 minutes was shortened.

• Journal of Korea Multimedia Society
• /
• v.23 no.12
• /
• pp.1447-1453
• /
• 2020

In this paper, we propose a method for no-reference quality assessment of sports videos using 3D shearlet transform and deep residual neural networks. In the proposed method, 3D shearlet transform-based spatiotemporal features are extracted from the overlapped video blocks and applied to logistic regression concatenated with a deep residual neural network based on a conditional video block-wise constraint to learn the spatiotemporal correlation and predict the quality score. Our evaluation reveals that the proposed method predicts the video quality with higher accuracy than the conventional no-reference video quality assessment methods.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2019.11a
• /
• pp.119-121
• /
• 2019

부가적 백색 잡음 모델(additive white Gaussian noise, AWGN에서 학습된 깊은 신경만 (deep neural networks)을 이용한 잡음 제거기는 제거하려는 잡음이 AWGN인 경우에는 뛰어난 성능을 보이지만 실제 카메라 잡음에 대해서 잡음 제거를 시도하였을 때는 성능이 크게 저하된다. 본 논문은 U-Net 구조의 깊은 인공신경망 모델에 residual block을 결합함으로서 실제 카메라 영상에서 기존 알고리즘보다 뛰어난 성능을 지니는 신경망을 제안하다. 제안한 방법을 통해 Darmstadt Noise Dataset에서 PSNR과 SSIM 모두 CBDNet 대비 향상됨을 확인하였다.