线列红外探测器的非均匀性会导致获取的红外图像含带固定噪声,这将严重影响图像的观测识别。提出一种基于双边滤波约束的局部恒定统计校正方法。该方法在恒定统计的非均匀校正方法的基础上,根据场景特性,对响应信号采用双边滤波方法进行滤波分析,并在统计结果中加以约束限制,可有效地解决原图像的非均匀性,并且避免了恒定统计算法中的“鬼影”与拖尾等问题。实验结果表明提出的算法校正效果明显,收敛速度快,易于工程化实现,可以大幅度提高线列红外成像系统的观测质量。 The inhomogeneity of line-array infrared detector will lead to the acquisition of infrared images with fixed noises, which will seriously affect the image recognition. A local constant statistical correction method based on bilateral filter constraints is proposed. Based on the constant statistical nonuniformity correction method, this method filters the response signals by using the bilateral filtering method according to the characteristics of the scene and restricts the statistical results, which can effectively solve the original image nonuniformity Avoid “ghosting” and smearing in constant statistics algorithms. The experimental results show that the proposed algorithm has obvious correction effect, fast convergence speed and easy engineering, which can greatly improve the observation quality of the line array infrared imaging system.