在传统红外图像非均匀性校正方法中，探测器均处在常温环境下，当探测器的环温在一个较宽的范围内变化时，这些方法的校正效果便会恶化、不适应。针对此问题，本文在原常温多点标定法的基础上提出了一种基于曲面拟合的校正方法，把原先标定曲线加上探测器环温维度扩展成标定曲面。随后的纵向对比实验和横向对比实验表明该方法能够显著降低非均匀性，并能适应环温变化的情况。

In the traditional infrared image non-uniformity correction method, detectors are at room temperature environment. When the environment temperature of detector is changing over a wide range, correction effect of these methods will be worse and not suited. To solve this problem, a non-uniformity correction method based on curved surfacefitting is proposed on the basis of original multiple points correction method. Fully considering the environment temperature of detector, original correction curve is increased to correction surface. Subsequent longitudinal comparative experiment and horizontal comparative experiment show that proposed method can greatly reduce the non-uniformity and adapt to the case of changing environment temperature.