Analysis And Modification On Imaging Aberration Caused By Light Oscillation In Turbulence

Videometrics is used to conduct survey by capturing the images of objects, and it is always carried out in atmospheric circumstance. Various motions of the air could lead to atmospheric turbulence, which always confines the upper precision limit of survey. This phenomenon has been noted and studied long time ago, while systematic research about it in videometrics could rarely be found. Firstly, some research progress and modification methods of the interaction between atmospheric turbulence and light propagation in some other related realm were summarized. Later, some fundamental physical properties, as well as some aero-optical effects, were presented. Thirdly, some basic survey experiments were introduced to find out what degree of affection could some factors exert on the precision. To modify the errors generated by aero-optical effects, a modification method of Kalman filter and a multi-symbol or multi-camera based imaging-constraint-method were both advanced. At last, several advices were given for further research.The experiments designed to illustrate the law of atmospheric disturbance give some results. On the one hand, the effects of disturbance will be more serious as long as the turbulence is fiercer, the measuring distance is farther, or the focused disturbance comes closer to the camera. On the other hand, different lens will exert different degree of disturbance on the measurement, and it is caused by several coupling factors.According to the spectrum character of turbulence, the Kalman filtering method was proposed to eliminate the unbiased component of disturbance. Inspired by the general method of mechanics, pseudo-kinetic equation of image point was set up based on center-penetrating imaging model. To enlarge its generality, the filtering model was developed into a new third-order one. The ability of following imaging point was approved by applying the third-order model to filter the coordinate of imaging point of kinetic symbol object. By comparing the coordinates of imaging points and the pose results of symbol object after filtering respectively to their original ones, the Kalman filtering method was proved to be quite effective to eliminate higher frequency component of disturbance, while useless to very low frequency disturbance.The formation and motion of shear turbulence within two moving plates was simulated by adopting CFD(Computational Fluid Dynamics) method. Based on the data of simulation, the deviation of light beam near a vortex was analyzed; the variation of the deviation, responding to motion of the flow field, was also analyzed. This work validated the effectiveness of the imaging-constraint-method. In order to realize this method, the imaging constraint equations of single-camera&multi-symbol and that of multi-camera&multi-symbol were both set up, and their solving methods were also presented. The corresponding multi-camera experiments adopted only two cameras for the sake of simplicity. The results of experiments unveiled that the two-camera group removed much more disturbance than the single-camera one. The reason leading to this phenomenon was that the “long periodical oscillation” deviates wholly the view field of either single camera.