• 제목/요약/키워드: matrix correction

검색결과 179건 처리시간 0.029초

GPU 기반 임베디드 시스템에서 대용량 데이터의 안정적 수신을 위한 ECC 연산의 가속화 (Acceleration of ECC Computation for Robust Massive Data Reception under GPU-based Embedded Systems)

  • 권지수;박대진
    • 한국정보통신학회논문지
    • /
    • 제24권7호
    • /
    • pp.956-962
    • /
    • 2020
  • 최근 임베디드 시스템에서 사용되는 데이터의 크기가 증가함에 따라, 대용량의 데이터를 안전하게 수신하기 위한 ECC (Error Correction Code) 복호화 연산의 필요성이 강조되고 있다. 본 논문에서는 GPU가 내장된 임베디드 시스템에서 해밍 코드를 사용하여 ECC 복호화를 할 때, 신드롬 벡터를 계산하는 연산의 수행을 가속할 방법을 제안한다. 제안하는 가속화 방법은, 복호화 연산의 행렬-벡터 곱셈이 희소 행렬을 나타내는 자료 구조 중 하나인 CSR (Compressed Sparse Row) 형식을 사용하고, GPU의 CUDA 커널에서 병렬적으로 수행되도록 한다. 본 논문에서는 GPU가 내장된 실제 임베디드 보드를 사용하여 제안하는 방법을 검증하였고, 결과는 GPU 기반으로 가속된 ECC 복호화 연산이 CPU만을 사용한 경우에 비하여 수행 시간이 감소하는 것을 보여준다.

비선형 감마 곡선 알고리즘 개선을 위한 구간 분할 다항식 곡선 접합 (The Segmented Polynomial Curve Fitting for Improving Non-linear Gamma Curve Algorithm)

  • 장정훈;조호상;장원우;강봉순
    • 융합신호처리학회논문지
    • /
    • 제12권3호
    • /
    • pp.163-168
    • /
    • 2011
  • 본 논문은 감마보정을 위한 비선형 곡선 알고리즘의 개선에 관한 연구이다. 기존의 비선형 감마 곡선 생성 방법은 Gauss-Jordan 역행렬을 적용한 최소 자승 다항식(Least Square Polynomial)을 사용하였다. 이 방법은 다항식 계수 값 계산 과정 중 고차행렬의 역행렬 연산에서 $10^{-11}$ 이하의 매우 작은 값은 절단함으로써 곡선접합의 정밀도가 감소된다. 또한 입력으로 사용되는 샘플 포인트가 10-bit 기준으로 0~1023의 밝기 값에 대하여 고루 분포되어있는 경우에만 정확한 동작이 가능하다. 본 논문은 이러한 기존 알고리즘의 단점을 보완하기 위하여, 고차 다항식의 계수 값을 반데몬드 행렬(Vandemond Matrix)에 SVD분해(Singular Value Decomposition)와 QR분해법(QR Decomposition)을 적용하여 행렬의 고유치와 직교성분만으로 연산하였다 또한, 입력 데이터의 구간을 분할하여 각 구간의 다항식을 생성하고, 새롭게 생성된 다항식을 이용하여 곡선 접합을 수행하도록 하였다. 입력 데이터와 곡선 접합결과의 평균제곱오차(Mean Square Error: MSE)와 표준편차(Standard Deviation: STD)를 통한 오차율 비교 결과 최하위 비트(Least Significant Bit: LSB) 에러 범위에서 MSE가 약 $10^{-9}$ 이고 STD는 약 $10^{-5}$로 정밀도가 향상되었다.

New Watermarking Technique Using Data Matrix and Encryption Keys

  • Kim, Il-Hwan;Kwon, Chang-Hee;Lee, Wang-Heon
    • Journal of Electrical Engineering and Technology
    • /
    • 제7권4호
    • /
    • pp.646-651
    • /
    • 2012
  • Meaningful logos or random sequences have been used in the current digital watermarking techniques of 2D bar code. The meaningful logos can not only be created by copyright holders based on their unique information, but are also very effective when representing their copyrights. The random sequences enhance the security of the watermark for verifying one's copyrights against intentional or unintentional attacks. In this paper, we propose a new watermarking technique taking advantage of Data Matrix as well as encryption keys. The Data Matrix not only recovers the original data by an error checking and correction algorithm, even when its high-density data storage and barcode are damaged, but also encrypts the copyright verification information by randomization of the barcode, including ownership keys. Furthermore, the encryption keys and the patterns are used to localize the watermark, and make the watermark robust against attacks, respectively. Through the comparison experiments of the copyright information extracted from the watermark, we can verify that the proposed method has good quality and is robust to various attacks, such as JPEG compression, filtering and resizing.

Two Messages out of One 2D Matrix Bar Code

  • Cvitic, Filip;Pavcevic, Mario Osvin;Pibernik, Jesenka
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • 제9권3호
    • /
    • pp.1105-1120
    • /
    • 2015
  • With the proposed principle of two-dimensional matrix bar code design based on masks, the whole surface of a 2D bar code is used for creating graphic patterns. Masks are a method of overlaying certain information with complete preservation of encoded information. In order to ensure suitable mask performance, it is essential to create a set of masks (mask folder) which are similar to each other. This ultimately allows additional error correction on the whole code level which is proven mathematically through an academic example of a QR code with a matrix of size $9{\times}9$. In order to create a mask folder, this article will investigate parameters based on Weber's law. With the parameters founded in the research, this principle shows how QR codes, or any other 2D bar code, can be designed to display two different messages. This ultimately enables a better description of a 2D bar code, which will improve users' visual recognition of 2D bar code purpose, and therefore users' greater enjoyment and involvement.

3 차원 곡면에 정밀 인쇄를 위한 공정 변수에 따른 이미지 보정에 관한 연구 (A study of correction dependent on process parameters for printing on 3D surface)

  • 송민섭;김효찬;이상호;양동열
    • 한국정밀공학회:학술대회논문집
    • /
    • 한국정밀공학회 2005년도 춘계학술대회 논문집
    • /
    • pp.749-752
    • /
    • 2005
  • In the industry, three-dimensional coloring has been needed for realistic prototype from rapid prototyping. Z-corporation developed a 3D printer which provides three-dimensional colored prototype. However, the existing process cannot be adopted to models from other rapid prototyping process. In addition, time and cost for manufacturing colored prototype still remain to be improved. In this study, a new coloring process using ink-jet head is proposed for color printing on three-dimensional prototype surface. Process parameters such as the angle and the distance between ink-jet nozzle and the three-dimensional surface should be investigated from experiments. The correction matrix according to sloped angle to minimize the distortion of 2D image was proposed by analysis of printing error. Therefore, approximated method for angle and discrete length according to the radius of curvature for printing on the curved surface was proposed. By printing image on the doubly curved surface, the method was verified. As a practical example, helmet was chosen for printing images on the curved surface. The character images were applied with approximated method for angle and discrete length and was printed on the helmet surface.

  • PDF

Characteristics of Magnetic Resonance-Based Attenuation Correction Map on Phantom Study in Positron Emission Tomography/Magnetic Resonance Imaging System

  • Hong, Cheolpyo
    • 한국의학물리학회지:의학물리
    • /
    • 제31권4호
    • /
    • pp.189-193
    • /
    • 2020
  • An MR-based attenuation correction (MRAC) map plays an important role in quantitative positron emission tomography (PET) image evaluation in PET/magnetic resonance imaging (MRI) systems. However, the MRAC map is affected by the magnetic field inhomogeneity of MRIs. This study aims to evaluate the characteristics of MRAC maps of physical phantoms on PET/MRI images. Phantom measurements were performed using the Siemens Biograph mMR. The modular type physical phantoms that provide assembly versatility for phantom construction were scanned in a four-channel Body Matrix coil. The MRAC map was generated using the two-point Dixon-based segmentation method for whole-body imaging. The modular phantoms were scanned in compact and non-compact assembly configurations. In addition, the phantoms were scanned repeatedly to generate MRAC maps. The acquired MRAC maps show differently assigned values for void areas. An incorrect assignment of a void area was shown on a locally compact space between phantoms. The assigned MRAC values were distorted using a wide field-of-view (FOV). The MRAC values also differed after repeated scans. However, the erroneous MRAC values appeared outside of phantom, except for a large FOV. The MRAC map of the phantom was affected by phantom configuration and the number of scans. A quantitative study using a phantom in a PET/MRI system should be performed after evaluation of the MRAC map characteristics.

A Symbiotic Evolutionary Design of Error-Correcting Code with Minimal Power Consumption

  • Lee, Hee-Sung;Kim, Eun-Tai
    • ETRI Journal
    • /
    • 제30권6호
    • /
    • pp.799-806
    • /
    • 2008
  • In this paper, a new design for an error correcting code (ECC) is proposed. The design is aimed to build an ECC circuitry with minimal power consumption. The genetic algorithm equipped with the symbiotic mechanism is used to design a power-efficient ECC which provides single-error correction and double-error detection (SEC-DED). We formulate the selection of the parity check matrix into a collection of individual and specialized optimization problems and propose a symbiotic evolution method to search for an ECC with minimal power consumption. Finally, we conduct simulations to demonstrate the effectiveness of the proposed method.

  • PDF

A Novel Voltage-Programming Pixel with Current-Correction Method for Large-Size and High-Resolution AMOLEDs on Poly-Si Backplane

  • In, Hai-Jung;Bae, Joon-Ho;Kang, Jin-Sung;Kwon, Oh-Kyong;Chung, Ho-Kyoon
    • 한국정보디스플레이학회:학술대회논문집
    • /
    • 한국정보디스플레이학회 2005년도 International Meeting on Information Displayvol.II
    • /
    • pp.901-904
    • /
    • 2005
  • A novel active matrix organic light diodes (AMOLEDs) voltage-programming pixel structure with current-correction method is proposed for largesize and high-resolution poly-Si AMOLED panel applications. The HSPICE simulation results shows that the maximum error of emission current in proposed pixel is 1.536%, 2.45%, and 2.97% with the ${\pm}12.5%$ mobility variation and ${\pm}0.3V$ threshold voltage variation for 30-, 40-, and 50-inch HDTV panels, respectively.

  • PDF

동적-외연적 유한요소법을 이용한 차체 판넬 성형해석에 관한 연구 (A Study of Auto-body Panel Correction of Forming Analysis that Use Dynamic-extensive Finite Element Method)

  • 정동원;황재신
    • 한국정밀공학회지
    • /
    • 제21권10호
    • /
    • pp.115-126
    • /
    • 2004
  • In the present work a finite element formulation using dynamic-explicit time integration scheme is used for numerical analysis of auto-body panel stamping processes. The lumping scheme is employed for the diagonal mass matrix and dynamic explicit formulation. Analyzed auto-body panel stomping process correction of forming using software called Dynaform using dynamic extensive method. Further, the simulated results for the auto-body panel stamping processes are shown and discussed. Its application is being increased especially in the automotive industrial area for the cost reduction, weight saving, and improvement of strength.

On Effects of Large-Deflected Beam Analysis by Iterative Transfer Matrix Approach

  • 신중호
    • 한국기계연구소 소보
    • /
    • 통권18호
    • /
    • pp.131-136
    • /
    • 1988
  • A small-deflected beam can be easily solved by assuming a linear system. But a large-deflected beam can not be solved by superposition of the displacements, because the system is nonlinear. The solutions for the large-deflection problems can not be obtained directly from elementary beam theory for linearized systems since the basic assumptions are no longer valid. Specifically, elementary theory neglects the square of the first derivative in the beam curvature formula and provides no correction for the shortening of the moment-arm cause by transverse deflection. These two effects must be considered to analyze the large deflection. Through the correction of deflected geometry and internal axial force, the proposed new approach is developed from the linearized beam theory. The solutions from the proposed approach are compared with exact solutions.

  • PDF