针对红外成像系统盲元检测中，传统窗口全局阈值法阈值选取的局限性，结合盲元响应随机性特点，提出了一种基于超像素分割的盲元检测算法，给出了分割区域像素数及检测过程阈值设置方法；针对盲元校正问题，提出了一种基于结构相似度和空间邻域距离加权的相关像素插值的盲元校正算法，研究了不同加权方法对估计误差的影响。最后通过实验验证了算法的准确性和有效性，结果表明：本文提出的算法检测结果准确率高、漏检率低、虚警率低，校正后图像的 RMSE低于邻域均值法（AN， Average Neighboring method）和最近邻替代（NN，Nearest Neighboring method）算法。

Aiming at the limitation of threshold selection of traditional window global threshold method in blind detection of infrared imaging system and combining with the random characteristics of blind response, a blind pixel detection algorithm based on super-pixel segmentation is proposed. A blind-element correction algorithm is proposed based on the structure similarity and spatial neighborhood distance-weighted pixel interpolation. The influence of different weighting methods on the estimation error is studied. Finally, the accuracy and validity of the algorithm are verified by experiments. The results demonstrate that the proposed algorithm has high accuracy, low missed detection rate, and low false alarm rate. Moreover, the RMSE of the corrected image is lower than that of the AN(Average Neighboring method) and NN(Nearest Neighboring method) algorithms.