这里研究了基于图像理解的无人机打击效果评估方法。重点阐述了轮廓提取技术、模板匹配技术和高层评估系统。该算法通过对打击前后图像进行变化检测,比较无人机打击前后的几何特征,参照不同部位不同的打击效果权重,对目标毁伤情况进行自动评估。实验表明,该算法提高了系统评估的精度,使评估结果更加准确。由于该算法具有通用性,该系统可以用于各类不同目标的打击效果评估。 Here we study the evaluation method of UAV combat effectiveness based on image comprehension. Focus on the outline of the extraction technology, template matching technology and high-level evaluation system. The algorithm detects the changes of the image before and after the attack, compares the geometric characteristics of the UAV before and after the attack, and refers to the weight of the attack effects in different parts to automatically evaluate the target damage. Experiments show that this algorithm improves the accuracy of system evaluation and makes the evaluation result more accurate. Due to the universality of the algorithm, the system can be used to evaluate the impact of various targets.