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。

Due to the non-uniformity of the linear array IR detector, there are noises available in the acquired IR images, which affect image’s observation and recognition seriously. A local invariable statistical correction method based on bilateral filter constraint is proposed, with the basis of invariable statistical non-uniform correction method, this method adopts bilateral filter for response signal for filtering analysis, and applies constrain limitation in the statistical result according to the scene characteristic. It resolves non-uniformity of original images efficiently as well as avoiding “ghosting” and tailing in existing algorithm. The experimental results indicate that the correction effect of this algorithm is prominent with fast convergence, and it is convenient for engineering implementation, this will enhance the observation quality of IR imaging system extensively.