红外焦平面阵列(IRFPA)的盲元既包括因材料与制造工艺的缺陷而导致的固定盲元,也包括因环境温度的漂移而出现的随机盲元。基于场景的盲元检测与补偿算法是去除这两种盲元,提高IRFPA成像质量的有效手段。针对目前滤波类场景检测算法无法有效区分弱小点目标和随机盲元的缺陷,重点研究了随机盲元的响应特性和噪声特性,并提出了一种基于模糊中值与时域累积的盲元自适应检测与补偿算法。首先利用模糊中值滤波器从场景中提取出潜在的盲元,并通过多帧累积确定固定盲元和随机盲元的正确分布,最后对盲元进行实时补偿。实验结果证明:该算法可以有效地实现对盲元的校正,同时避免对弱小点目标的误判别。 Blind elements of an infrared focal plane array (IRFPA) include both fixed blind elements due to defects in materials and manufacturing processes, as well as random blind elements due to ambient temperature drift. Scene-based blind detection and compensation algorithm is to remove these two blind elements and improve IRFPA imaging an effective means. Aiming at the defect that the filter class scene detection algorithm can not effectively distinguish the weak point and the random blind element, the response characteristic and the noise characteristic of the random blind element are studied emphatically, and a novel blind element based on fuzzy median and time-domain self-accumulation Adaptive detection and compensation algorithms. Firstly, the fuzzy median filter is used to extract the potential blind elements from the scene, and the correct distribution of the fixed blind elements and the random blind elements is determined by accumulating multiple frames. Finally, the blind elements are compensated in real time. Experimental results show that this algorithm can effectively correct blind pixels and avoid misidentification of weak dot targets.