Target strike result self-evaluation method based on image sequence analysis

The invention discloses a target strike result self-evaluation method based on seeker image sequence analysis, and belongs to the technical field of strike result evaluation. According to the method, for an image sequence collected by a seeker, target position information in each frame of image is obtained through target recognition algorithm processing, and a two-dimensional Gaussian model is adopted to predict a missile drop point position estimation value; and further comparing the estimated value of the drop point position with the actual target position to realize the preliminary evaluation of the strike result. The problems that images are difficult to obtain after striking, and real-time evaluation cannot be supported are solved, and support is provided for on-board distribution of striking targets. 本发明公开一种基于导引头图像序列分析的目标打击结果自评估方法，属于打击结果评估技术领域。本发明针对导引头采集的图像序列，通过目标识别算法处理，获取每一帧图像中的目标位置信息，并采用二维高斯模型预测导弹落点位置估计值；进一步将落点位置估计值与实际目标位置相比较，实现对打击结果的初步评估。本发明解决了打击后图像获取难、无法支撑实时评估等困难，为打击目标临机分配提供支撑。