Study On The Key Technologies Of Online Micro-Vision Measurement System

As the key part of Inertial Confinement Fusion, frequently referred as ICF, micro-target assembly precision directly affects the success rate of the ignition experiment. Micro-target assembly for inertial fusion is being investigated and has gained some achievement by a number of institutions throughout the world. Reports of the online detection of the relative position among components of micro-target have been rarely found in our country. After the analysis of the structural features, the geometric parameters and the relative position of the components of the assembly, an online measurement system, which base on micro-vison, has been proposed in this paper.An orthogonal micro-imaging system is used to inspect the assemble process of micro-target. To avoid fabricating especially calibrating object needed in general calibration algorithms, a camera calibration method using standard ball is proposed for micro-image measuring system. The method simplifies the computational modeling and reduces calibration parameters. Using the feature that the contours of standard ball placed in the common field of view of cameras are always visible from any position, multiple cameras in micro-vision system can be calibrated simultaneously, so the process of calibration will be simplified. The experimental results show that calibration precision is high.A new image definition evaluation function based on image moments is proposed on account of analyzing the image-forming principle of out-of-focus image, and its definition is presented. New evaluation function is composed by the weighting of image moments of different order, and the sensitivity and noise immunity of evaluation function can be flexibly controlled by adjusting weighting coefficient. Experiment results show that under low contrast, as compared to existing image definition evaluation functions, which are inclined to lose their ideal characteristics, the image resolution evaluation function based on image moments not only effectively maintained ideal curve characteristics such as unimodality and unbiasedness, but also had obvious advantages in the aspects of sensitivity and noise immunity.The imaging processing algorithms for the computer vision based measurement systems are studied. The preprocessing and threshold segmentation algorithms are investigated. Through the analysis of common edge detection algorithms and the simple edge, get the target’s accurate and complete edge with low noise. This essay deeply researches the detection of lines, circles and ellipses in the image. Especially the methods for the detection of circles and ellipses have more improvement and new theories than the essays before, reducing much calculation and time.