为提高搜救航拍中生成大视场红外图像的效率，提出一种序列图像提取关键帧的方法。分析图像获取系统的运动参数和获取红外图像的内方位元素，采用（.,.,.）转角系统作为图像的外方位元素，根据摄影测量中的共线方程式计算序列图像间的重叠率，并以相位相关法对计算出的重叠率进行验证。最后根据计算出的重叠率和采用的 Harris拼接算法所要求的最佳重叠率，来确定序列图像中的关键帧。试验结果表明，计算出的重叠率可以作为关键帧提取的标准，待拼接的图像数量减少，生成大视场图像的效率提高 88.71%。

To improve the efficiency of wide field infrared image generation in research and rescue aerial photography， a method for key frame extraction from image sequence is proposed in this paper. Firstly, the motion parameters of image acquiring system and image’s elements of exterior and interior orientation are considered. Secondly, the（φ，ω，κ）angle system is adopted and the overlap rate between images is calculated according to the collinear equation. Finally, the overlap rate calculated and the best overlap rate offered by Harris algorithm are consider to determine the key frames. The result shows the overlap rate which is calculated can be indicator for key frame extraction, and the efficiency of wide field image is promoted to 88.71%.