基于特征分解的红外焦平面非均匀性校正算法

针对目前红外焦平面自适应场景校正算法工程应用的局限，提出了一种基于红外焦平面非均匀性特征分解的场景校正算法。分析了红外焦平面非性匀性构成因素，把其中的高频部分分解成盲点、斑块、行列非均匀性等，把缓慢变化的低频部分分解成梯度渐变非均匀性；分别对各类非均匀性采用不同的校正算法；合并校正结果，得到校正后的图像。实验结果表明，该算法校正精度高、收敛速度快、抑止目标退化能力强，适合工程应用。

Non-uniformity correction algorithm based on image decomposition

A nonuniformity correction algorithm based on image decomposition of nonuniformity in IRFPA was proposed, for the engineering application limitations of traditional adaptive nonuniformity for IRFPA. After analysis of the constructing factor of nonuniformity, the algorithm decomposes the high-frequency part of nonuniformity into bright pixels, row and column nonunniformity, and the low-frequency part of nonuniformity into gradient nonunifority, and then corrects every kind of nonunifority using a special algorithm, and completes the correction by combination. The algorithm was tested in a infrared imaging system and compared with other adaptive non-uniformity correction algorithms. According to the analysis of simulation results, this algorithm has the advantage of strong ability to eliminate the target fade-out, fast convergence speed and high precision.