基于黑体的自适应选点多点非均匀性校正

Adaptive Point Selection Multipoint Nonuniformity Correction Based on Black Body

红外成像受探测器材料工艺的不足、读出电路的非均匀性、温度和光学结构的影响,具有非均匀性,原始的红外图像需要进行非均匀性校正。非均匀性校正一般分为基于定标和基于场景两种,其中基于定标的两点校正得到了广泛的应用。本文基于现有的基于定标的校正算法,分析探测器响应曲线,提出一种基于三阶多点校正算法,既能更好的拟合探测器相应曲线,又能更好的校正离散点。对于所需要定标点,与待校正图像进行自适应的对比选点,选出待校正图像温度附近效果最好的点,最终计算量大大降低,且精度和准确率得到了一定的提升。

Infrared imaging has nonuniformity due to inadequate detector material processing,nonuniformity of readout circuit,temperature and optical structure.The original infrared image needs to be corrected for nonuniformity.Non-uniformity correction is generally divided into two types:calibration-based and scene-based.Two-point calibration based on calibration is widely used.In this paper,based on the existing calibration-based calibration algorithm,the response curve of the detector is analyzed,and a three-dimensional multipoint calibration algorithm is proposed,which can not only better fit the corresponding curve of the detector,but also better correct the discrete points.For the desired punctuation point,make an adaptive comparison with the image to be corrected,and select the best point near the temperature of the image to be corrected for comparison.The result shows that the computational load is greatly reduced,and the accuracy and accuracy are improved.