Fringe Projection Profilometry System Calibration and Multiview Range Image Registration

• Main Authors: Shin-Yuan Sue, 蘇鑫元
• Other Authors: 李吉群
• Format: Others
• Language: zh-TW
• Published: 2006
• Online Access: http://ndltd.ncl.edu.tw/handle/96044163737057064783

碩士 === 中興大學 === 機械工程學系所 === 94 === 　The CMM is used for surface profilometry traditionally, but it is not only expensive but also time consuming. With the computer technique promotes, the measurement uses the computer vision became convenient and inexpensive. The purpose of this paper is to establish a three-dimensional contours rebuilding system, which is economical and simple. 　This paper describes the usage the liquid crystal display (LCD) which projects the fringe projection onto the object’s surface. Then the triangulation principle is used to get the surface profile. For improving the accuracy of the system, we have to modify the system error, which results from lens distortion and the nonlinearity of the VGA display card, LCD projector and image grab. Generally speaking, there is always a Gamma correction for display equipment to accommodate the human’s visual system, LCD for example. This correction makes brightness nonlinear problem and influences the phase accuracy, so we need calibration to improve the brightness nonlinear problem. The measurement system obtains the single view profile information by means of the Gray-code method and phase-shifting technique to calculate the absolute phase, and the mapping relationship between the measured points and image pixel. Then we execute the range image registration and shortest path algorithm for all views. According to the shortest path and the transformation of each view, we can acquire the initial model. Finally, we execute the simultaneous regulating produce to obtain the whole contours of component.