• Proceedings of the Korea Information Processing Society Conference
• /
• 2009.04a
• /
• pp.90-93
• /
• 2009

본 논문은 자동차 자동운행 시스템 연구 분야의 한 부분인 자동차 운행 중 도로상에 위치한 교통 신호등을 추적을 통해 검출하고, 인식하기 위한 방법과 관련된 연구이다. 교통 신호등은 색상 정보를 포함한 광원을 갖는 물체로서 표현되어지고 운전자에게 안전을 위해 준수해야 할 신호정보로써 제공되어 진다. 본 논문에서는 이러한 교통신호등의 인식을 위해 명도 분포도를 이용하여 관심영역을 필터링하고, 마스크와 HSI 색 공간영역에서의 색상과 채도, 밝기 정보를 이용한 유효값을 검출, 좌표변환, 보간법, YUV 모델을 이용한 그레이 영상으로의 변환, 닫힘 연산, 선명화 연산, 템플릿 매칭 방법을 적용함으로써 가로등과 같은 주변 환경이 갖는 색정보로부터 교통 신호등의 신호를 검출하고 인식하도록 하였다.

• Journal of Broadcast Engineering
• /
• v.25 no.3
• /
• pp.325-337
• /
• 2020

Versatile Video Coding (VVC) is a new video coding standard that is being developed by the Joint Video Experts Team of ISO/IEC/ITU-T and it has adopted various technologies including screen content coding tools. Screen contents have a feature that blocks are likely to have diagonal edges like character regions. If triangular partitioning coding is allowed for screen contents having such the feature, coding efficiency would increase. This paper proposes a intra prediction method using triangular partitioning prediction for screen content coding. Similar to the Triangular Prediction Mode of VVC that supports the triangular partitioning prediction, the proposed method derives two prediction blocks using Horizontal and Vertical modes and then it blends the predicted blocks applying masks with triangle shape to generate a final prediction block. The experimental results of the proposed method showed an average of 1.86%, 1.49%, and 1.55% coding efficiency in YUV, respectively, for VVC screen content test sequences.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.20 no.6
• /
• pp.786-792
• /
• 2010

This paper presents a real-time lane detection method which can accurately find the lane-mark boundaries in complex road environment. Unlike many existing methods that pay much attention on the post-processing stage to fit lane-mark position among a great deal of outliers, the proposed method aims at removing those outliers as much as possible at feature extraction stage, so that the searching space at post-processing stage can be greatly reduced. To achieve this goal, a grid-based morphology operation is firstly used to generate the regions of interest (ROI) dynamically, in which a directional edge-linking algorithm with directional edge-gap closing is proposed to link edge-pixels into edge-links which lie in the valid directions, these directional edge-links are then grouped into pairs by checking the valid lane-mark width at certain height of the image. Finally, lane-mark colors are checked inside edge-link pairs in the YUV color space, and lane-mark types are estimated employing a Bayesian probability model. Experimental results show that the proposed method is effective in identifying lane-mark edges among heavy clutter edges in complex road environment, and the whole algorithm can achieve an accuracy rate around 92% at an average speed of 10ms/frame at the image size of $320{\times}240$.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2006.11a
• /
• pp.73-76
• /
• 2006

H.264는 MPEG-2, MPEG-4 Part 2, H.263 등 기존의 비디오 압축 표준들에 비해 우수한 화질과 부호화 효율을 제공하여 차세대 비디오 압축 표준으로서 널리 사용될 전망이다. 현재 H.264표준화 그룹인 JVT에서는 디지털 시네마 영상을 위한 Advanced 4:4:4 프로파일에 대한 표준화가 진행 중이다. 이 프로파일은 기존의 프로파일과 달리 화소당 8-12비트의 영상을 지원하며, YUV 영상대신 RGB 영상을 입력영상으로 사용한다. 디지털 시네마 영상은 보통 HD급 이상의 화면 크기를 가지며 초당 24Hz의 프레임율을 가진다. 이러한 영상에서는 화소간의 공간적 유사성이 매우 높아지는 경향이 있으며, 30Hz 영상에 비해 시간적 유사성이 감소하는 경향이 있다. 그 결과 H.264로 디지털 시네마 영상 압축 시, 공간 예측을 통한 Intra 매크로블록의 비율이 다른 테스트 영상들에 비해 월등히 높다는 것을 알 수 있다. 이는 디지털 시네마 영상 압축 시, 화면간 예측을 수행하는 ME/MC에 비해 공간 예측이 효율적이다는 것을 입증한다. 화면간 예측의 성능을 향상하기 위해 본 논문에서는 ME/MC 후 생성된 잔차 신호들을 간단히 변환하는 방법을 제안한다. 간단한 변환 기술이 추가되어 화면간 예측의 압축 성능이 향상됨은 물론, Inter 프레임에서 화면간 예측과 공간예측을 모두 사용하였을 때 전체적인 압축성능이 향상함을 실험을 통하여 확인 할 수 있었다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2008.11a
• /
• pp.141-144
• /
• 2008

IPTV, VOD와 같은 스트리밍 서비스와 유료 케이블 TV 방송 채널, 유료 위성방송 채널 등에는 반드시 제한 수신 시스템(Conditional Access System, CAS)이 필요하다. CAS시스템은 인증 받은 사용자에게는 깨끗한 화면을 보여주어야 하고 인증 받지 않은 사용자에게는 정상적으로 화면을 즐길 수 없도록 하여야 한다. 이러한 환경을 만들기 위하여 추가비용이 적은 알고리즘이 필요한데 그 방법으로 암호화를 제안한다. 본 논문에서는 CAS 등의 인증시스템을 위하여 H.264 영상의 선택적 암호화를 구현하였다. 제안하는 방법은 여러 가지 포맷으로 된 콘텐츠를 트랜스코딩하여 YUV 형식으로 변환한 뒤, 이것을 H.264 코덱을 이용하여 압축한 다음 필요에 따라 특정한 부분을 암호화하는 것이다. 실험을 통하여 암호화 키가 없는 미 인증 사용자는 영상을 제대로 볼 수 없었고 암호화의 강도를 강하게 할 경우 재생은 되지만 영상의 확인이 불가능함을 확인하였다. 또한 300프레임의 영상을 암호화 하는데 평균 71.3초가 걸려 속도 역시 빠름을 확인하였다. 제안하는 내용은 IPTV, VOD와 같은 스트리밍 서비스에서의 사용자 인증 및 저작권 보호 등의 분야에 유용하게 사용될 것이다.

• Journal of the Institute of Convergence Signal Processing
• /
• v.9 no.4
• /
• pp.245-252
• /
• 2008

In this paper, we propose a new method for improving the H 264 encoding process and motion estimation part. Our approach is a method to reduce the encoding running time through the omission of reference frame in the mode selection process of H 264 and an improvement of SAD computing process. To evaluate the proposed method, we used the H 264 standard image of QCIF size and TIN 4:2:0 format. Experimental results show that proposed SAD algorithm 1 can improve the speed of encoding runnung time by an average of 4.7% with a negligible degradation of PSNR. However, SAD algorithm 2 can improve the speed of encoding runnung time by an average of 9.6% with 0.98dB degradation of PSNR.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.10C
• /
• pp.1414-1424
• /
• 2004

In this paper, an algorithm using wavelet transform for detecting a cut that is a radical scene transition point, and fade and dissolve that are gradual scene transition points is proposed. The conventional methods using wavelet transform for this purpose is using features in both spatial and frequency domain. But in the proposed algorithm, the color space of an input image is converted to YUV and then luminance component Y is transformed in frequency domain using 2-level lifting. Then, the histogram of only low frequency subband that may contain some spatial domain features is compared with the previous one. Edges obtained from other higher bands can be divided into global, semi-global and local regions and the histogram of each edge region is compared. The experimental results show the performance improvement of about 17% in recall and 18% in precision and also show a good performance in fade and dissolve detection.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.11C
• /
• pp.1518-1526
• /
• 2004

In this paper, an algorithm using wavelet transform for detecting a cut that is a radical scene transition point, and fade and dissolve that are gradual scene transition points is proposed. The conventional methods using wavelet transform for this purpose is using features in both spatial and frequency domain. But in the proposed algorithm, the color space of an input image is converted to YUV and then luminance component Y is transformed in frequency domain using 2-level lifting. Then, the histogram of only low frequency subband that may contain some spatial domain features is compared with the previous one. Edges obtained from other higher bands can be divided into global, semi-global and local regions and the histogram of each edge region is compared. The experimental results show the performance improvement of about 17% in recall and 18% in precision and also show a good performance in fade and dissolve detection.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.47 no.5
• /
• pp.50-60
• /
• 2010

For the last half a century, the personal computer and software industries have been prosperous due to the incessant evolution of computer systems. In the 21st century, the embedded system market has greatly increased as the market shifted to the mobile gadget field. While a lot of multimedia gadgets such as mobile phone, navigation system, PMP, etc. are pouring into the market, most industrial control systems still rely on 8-bit micro-controllers and simple application software techniques. Unfortunately, the technological barrier which requires additional investment and higher quality manpower to overcome, and the business risks which come from the uncertainty of the market growth and the competitiveness of the resulting products have prevented the companies in the industry from taking advantage of such fancy technologies. However, high performance, low-power and low-cost hardware and software platforms will enable their high-technology products to be developed and recognized by potential clients in the future. This paper presents such a platform for industrial embedded systems. The platform was designed based on Telechips TCC8300 multimedia processor which embedded a variety of parallel hardware for the implementation of multimedia functions. And open-source Embedded Linux, TinyX and GTK+ are used for implementation of GUI to minimize technology costs. In order to estimate the expected performance and power consumption, the performance improvement and the power consumption due to each of enabled hardware sub-systems including YUV2RGB frame converter are measured. An analytic model was devised to check the feasibility of a new application and trade off its performance and power consumption. The validity of the model has been confirmed by implementing a real target system. The cost can be further mitigated by using the hardware parts which are being used for mass production products mostly in the cell-phone market.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.53 no.5
• /
• pp.143-151
• /
• 2016

The aim of this study is to investigate the optimal algorithm to extract medical radiation induced pixel signal from complementary metal-oxide semiconductor (CMOS) sensors of smartphones camera. The pixel intensity and pixel number of smartphone camera were measured as the X-ray dose was increased. The front camera of the smartphone camera has low noise property and excellent dose response as compared to the back camera of the smartphone. The indirect method which uses scintillation crystal in front of the smartphone camera, couldn't improve the X-ray detection efficiency as compared to the direct method which does not use any scintillator in front of the smartphone camera. When we used the algorithm which employing threshold level on the pixel intensity and pixel number, the dose linearity was more higher for the pixel intensity rather for the pixel number. The use of pixel intensity of Y color component which represents the grey scale, would be efficient in terms of the radiation detection efficiency and reducing the complexity of the image processing. We expect that the radiation dose monitoring can be managed effectively and systematically by using the proposed radiation detection algorithm, thus eventually will contribute to the public healthcare.