The Study On Auto-focus Method For Optical Imaging System Based On Image Processing

It is a necessary condition for the high precision measurement to obtain definite images in the range photogrammetry systems. At present, all systems used in the ranges take a manual focusing method to get definite images, which does not work effectively and can be disturbed easily by various arbitrary factors, and with this method, it's hard to get definite images in the whole process of the measurement. The purpose of this paper is to propose a fast auto-focusing method based on image processing technology which can cope with the fast-moving targets with complicated shape in the range.After the principle and categorizations of auto-focusing method being introduced systematically, the advantage of the method based on image processing is analyzed in the paper, which obtains the definite images by driving the motorized lens to the optimization focus position using the information of the processed image as the focusing criterions. Firstly, the image definition functions are discussed. Then, with the comparison and analysis to all criterion functions, an appropriate image definition function for the auto-focusing system satisfied our study is chosen according to the character of target imaging after a series of experiments. Using the existing equipments of the laboratory, the auto-focus system is built, and the software of the system is written by the Viaual C++ programme language. To evaluate the performance of the whole auto-focusing system, experiments indoor and outdoor are both designed, and the results of the experiments are analyzed, which proved the feasibility of the auto-focus method for the moving targets proposed in this paper.The video-metric system for long distance railway geometry parameters measurement is one work of the research group, and it has the auto-focus function. This paper takes part in the amelioration and experiments work of the video-metric system. The results of the experiments are analyzed. The influence of the atmosphere turbulence to the auto-focus system is analyzed by experiments. To eliminate the negative influence of the turbulence, a method using curve-fitting method is proposed, and the experiments show that this method works well.