Font Size: a A A

Research On Key Technologies Of Laser Beam Quality Measurement

Posted on:2014-07-21Degree:MasterType:Thesis
Country:ChinaCandidate:X B ZhuFull Text:PDF
GTID:2268330422464822Subject:Optical Engineering
Abstract/Summary:PDF Full Text Request
This thesis uses CMOS camera as the beam sensor, analysis of the transformation whenthe beam go through a complex optical system, and discusses the issues of using a neutraldensity filter to do the beam attenuation, as well as the method of suppressing random noiseintroduced by CMOS Camera. Averaging the multi-pictures in continuous sampling andgetting the mean of single view by the method of neighbour average smoothing caneffectively improve the SNR. For spot images obtained from the underlying of PC system,the thesis introduces series of digital image processing, proposes a improved method thatcombine image segmentation and gravity calculating to locate spot center. This method canprevent the weight shift problem caused by uneven or partial gaps of the spot image. Thethesis also discusses of the method of identifying spot boundary. Using a gradient methodcan only judge visual boundaries of the spot, while using the method of tracking graytransitions, you can get the true boundaries. This thesis also presents the different sizecalibration methods, one is direct measurement when laser go through the surface of CMOStarget, the other is indirect measurement when equipping lens to the front of CMOS camera.This thesis studies the GNL iterative fitting algorithm of curve fitting for Gaussian beam,which are Gaussian curve fitting for gray curve and hyperbolic fitting for parameters (waist,divergence angle, M2factor, etc.)calculations. Applying the initial value, which calculated bythe parametric linear least square method, to the GNL iteration method, can ensure iterationconvergence and reduce the fitting errors of GNL iterative to a million.In this thesis, for supporting software of beam detection system, it completed5UMLdesign categories of class family as follow: image processing class, spot parameter class,matrix operations class, curve fitting class, and laser parameters class. The use of decorativepatterns to organize series of image processing, makes the software flexible and scalable.The thesis also shows a detection system (including hardware platforms and softwareapplications) which was built based on the research all above in2occasions, one is the1080nm laser mode observation, the other is homogenized piece features observed inexperimental situations, Both of them illustrate the flexibility and utility of the system.
Keywords/Search Tags:Beam quality, Curve fitting, Digital Image processing, Decorative patterns, C++programming
PDF Full Text Request
Related items