基于梯度场景的非均匀校正方法  

Nonuniformity Correction Method Based on Gradient Scenes

作　　者：周达标 王德江[1,3] 霍丽君[2,3] 刘让[1,2,3] 贾平[1,3] 

机构地区：[1]中国科学院长春光学精密机械与物理研究所中国科学院航空光学成像与测量重点实验室,吉林长春130033 [2]中国科学院大学,北京100049 [3]中国科学院长春光学精密机械与物理研究所,吉林长春130033

出　　处：《光学学报》2017年第5期338-347,共10页Acta Optica Sinica

基　　金：国家自然科学基金(61308099)

摘　　要：长波红外探测器经常被用于机载红外预警系统中,常受严重的非均匀性噪声干扰。为了校正探测器的非均匀性,补偿辐射响应非线性,提出了一种基于梯度场景的非均匀性校正方法。给出了探测器辐射响应非均匀性的观测模型;以标准黑体和梯度场景作为参考源,在理论上推导出校正系数表达式;利用原理样机进行了外场实验,并探测民航客机目标。实验结果表明:与基于黑体的两点校正方法相比,利用本文方法进行非均匀性校正后的图像,局部标准差峰值由8.57降低到2.39;对于相距50.64km的空中客车A319型客机,目标的信杂比由4.87提高到11.22。本文算法可以有效降低图像局部标准差,适用于机载红外预警系统。Long wave infrared detectors are often employed in airborne infrared early warning systems, which are often contaminated the with heavy nonuniformity noises. To correct the nonuniformity and compensate the nonlinearity of detectors, a nonuniformity correction method is proposed. The observation model of radiance nonuniformity is introduced. The correction coefficients are deduced in theory with the reference of a blackbody and a gradient scene. And, an outfield experiment with a proof-of-concept camera is performed to detect the airliners. Experimental results show that the peak value of local standard deviation of the corrected image decreases from 8.57 to 2.39, compared with the two-point correction method based on a blackbody. For the target of Airbus A319 from 50.64 km away, the signal-to-clutter ratio rises from 4.87 to 11.22. The proposed method can decrease the local standard deviation effectively, which has a significant contribution to airborne infrared early warning systems.

关 键 词：遥感 非均匀性校正 梯度场景 两点法 标定 预警 

分 类 号：TP751.1[自动化与计算机技术—检测技术与自动化装置]