针对非合作大目标的被动测距问题，介绍了一种基于单目图像序列和成像方向的测距系统，并进行了性能分析.这种测距系统的特点是无需初始距离导引，它对目标距离的估计借助于专门的定位方程，方程由目标特征线度、摄像机的空间坐标、目标在采样时刻对相机的方位角和俯仰角决定.其中，方位角和俯仰角通过光电经纬仪获得,摄像机的空间坐标通过GPS获得,目标特征线度通过相邻帧图像匹配得到.分析表明，该系统的性能主要取决于特征线度的准确度，其他测量误差如目标的方位角、俯仰角和摄像机的空间坐标对其影响较小.目标特征线度选取时,在相邻帧目标图像匹配点中，筛选出三个具有较大极限的尺度不变特征变换关键点，构造三角形及其外接圆，在外接圆上取经过三角形重心且平行或垂直于尺度不变特征变换主方向的弦线作为特征线度，这种特征线度对目标旋转不敏感.实验结果表明这种测距方法能够实现对非合作大目标的被动测距.

Aiming at the passive ranging to a noncooperate target, a system based on image sequence from single lens and the respective imaging direction is introduced with its performance. This ranging system does not need initial distance guiding. The distance from target to camera can be estimated by solving a special ranging equation, which is made from target′s characteristic linearity, camera′s coordinates in space, and the target′s azimuth and pitching to the camera. Among these data, azimuth and pitching are obtained by an electrooptical altazimuth. Camera′s coordinates come from a GPS, and the target′s characteristic linearity will be from the image matching between adjacent frames in the image sequence. Analysis shows that the ranging error is mainly determined by the precision of the characteristic linearity, other influencing factors have less impact, which include the measurement error of the target′s azimuth or pitch and the error of the camera′ coordinates. Thus a linearity can be obtained by selecting three matched points in each adjacent frames. These three points have strong Scale Invariant Feature Transform characteristic, with the points a triangle and its circumcircle can be settled. The chord across barycenter of the triangle can be taken as the best characteristic linearity. The experimental result indicates that this ranging method can realize the passive ranging to a noncooperate big target.