应用图像法对长焦距高分辨率航空相机实时检调焦过程中，由于外界环境因素的影响使得系统产生像移，调焦精度降低。当前对于由相对运动引起的像移和调焦精度两者的之间关系没有通用性的理论和确切的研究。从理论上分析了高分辨率航空图像线性像移和抖动像移的模型，给出了像移模型对评价函数的影响，得到像移与调焦曲线极大值呈负相关的关系。图像的像移量越大，调焦函数曲线的最大值下降越显著。利用二次函数拟合的方法获得离散评价函数点的拟合曲线，曲线最高点即对应调焦的最优对焦位置。搭建了动基座检调焦实验平台，实验结果表明当平台引起的图像像移量由1 pixel 增加至5 pixel 时，评价函数曲线的最大值由95.95%下降到87.64%。应用拟合曲线的方法得到的最佳对焦位置在系统半焦深范围之内，满足高分辨率航空相机调焦精度的要求，同时证明了像移理论分析的正确性。

In the use of real-time auto-focusing technology based on image processing of long range and high resolution aerial camera, the precision decreases due to the influence of image motion. Currently there is no universal theory and exact conclusion on the relationship between image motion and auto-focusing precision. The models of linear motion and jitter are analyzed theoretically, and the mathematical formulas illustrated the influence of image motion on auto-focusing measures are given. The image motion has a negative correlation to maximum value of focusing curve, the larger amounts of image motion, the smaller maximum value of the auto-focus measures. The best focusing position is located at the top of the fitting curve, by making use of the quadratic fitting technology to get auto-focusing curve. The motion base platform for auto-focusing is designed and set up. Experimental results show that when the image motion increases from 1 pixel to 5 pixel, the maximum value of curve decreases from 95.95% to 87.64% . At the same time, the best auto-focusing position on fitting curve is in the high resolution aerial camera’s half depth of focus. This result proves the correctness of image motion theoretical analysis.