• Title/Summary/Keyword: Image Processign

Search Result 4, Processing Time 0.024 seconds

Three-dimensional measurement of object surface and moving particles using at TV camera

  • Kawasue, Kikuhito;Iwamoto, Isamu;Ishimatsu, Takakazu
    • 제어로봇시스템학회:학술대회논문집
    • /
    • 1997.10a
    • /
    • pp.1302-1305
    • /
    • 1997
  • A new approach to the three-dimensional measurement of the object surface and moving particles is introduced. A single TV camera with an apparatus to add the circular bias to the image enables us to record the three-dimensional information of measuring points as streaks on a single image. Every shaped of the streak on the image plane is related to the position of the measuring point. the information is extracted form the image using an image processign technique.

  • PDF

A Study on Development of Velocity Measuring Method by PTV-AOM (PTV-AOM을 이용한 속도계측 기법의 개발에 관한 연구)

  • 조대환;김진구;이영호
    • Journal of Advanced Marine Engineering and Technology
    • /
    • v.21 no.5
    • /
    • pp.591-597
    • /
    • 1997
  • Instantaneous velocity distribution within coaxial circular pipe for measurement velocity of high speed is acquired simultaneously by applying the 2-dimensional PTV system consisting of a pulse generator(AOM:Acousto-Optical Modulator), a continuous-output laser and a PC image grabber together with experimental apparatus. The basic mechanism of AOM and vector identification method and performance-related image processing techniques are discussed. Representative measuring regions $90{\times}90$mm are selected and instantaneous vectors are represented and fully developed turbulent flow of maximum velocity up to about 1.0 m/sec is obtained.

  • PDF

Evaluationof Growing Crack-Tip Singularity in A533B Steel by Image Processing Technique (화상처리법을 이용한 A533B강의 진전균열특이장 평가)

  • Pyo, Chang-Ryul;Kim, Yeong-Jin
    • Transactions of the Korean Society of Mechanical Engineers A
    • /
    • v.21 no.1
    • /
    • pp.124-132
    • /
    • 1997
  • This paper describes an experimental and numerical study on growing ductile crack-tip behaviors. The hybrid experimental and numerical method by means of a computer image processign technique, was applied to the analysis of both base metal and weld metal CT specimens. In the weld metal specimen, the initial crack-tip was placed in front of fusion line, and the crack orientation was perpendicular to it. Finite element analysis of crack growth behaviors in both base and weld matal specimens made of A533B Class 1 steel were also performed to examine the effects of weldment on near crack-tip fields. a series of experimental studies on crack-tip behaviors have clearly shown the qualitative effects of material properties, especially a hardening exponent. The experimental and numerical results have also shown that weldment does not affect displacement and strain fields near a crack-tip while a stress field is influenced by the difference between yield stresses of both base and weld metals.

Implementation of High-Speed Fresnelet Transform using Daubechies's Filter (드뷔시 필터를 이용한 고속 프레넬릿 변환의 구현)

  • Seo, Young-Ho;Lee, Yoon-Hyuk;Kim, Dong-Wook
    • Journal of the Korea Institute of Information and Communication Engineering
    • /
    • v.21 no.4
    • /
    • pp.820-828
    • /
    • 2017
  • Service of digital hologram that has been recognized as a visual system for next generation requires various signal processing technologies. A transform is the most frequently used tool among signal processing techniques for 2-dimensional(D) and 3-D natural picture. A digital hologram has totally different property with a natural picture, so it is rarely efficient to apply transform tools used in 2-D image processing to a digital hologram. To overcome this a Fresnelet transform for a digital hologram has been proposed. We derive a Fresnelet transform by using the Daubechie's filter after applying an unitary Fresnel transform to a wavelet basis function. We also implement the transform as types of device and kernel code to improve operational performance. In consideration of the average time that is required for a pixel we can have observed the performance is improved up to 242 and 30 times for using the (9,7) and (5,3) filters in case of using device code.