论文部分内容阅读
针对红外焦平面阵列探测单元响应的非线性对非均匀性校正的影响,利用探测器响应的非线性模型,对卡尔曼滤波非均匀性校正算法进行扩展和改进,以有效地克服响应非线性对校正精度的影响。该算法先对图像进行非线性压缩,转换为线性图像,再利用线性模型下的卡尔曼滤波算法实施非均匀校正,然后对其进行取指数操作,即得到原图非均匀校正后的图像。实验结果表明,该算法不仅继承了原算法利用场景信息来最优地更新校正参数的估计,解决了探测器偏置和增益漂移对校正影响,而且还在一定程度上解决了响应非线性对校正性能的影响,从而获得了较好的非均匀性校正效果。
Aiming at the influence of non-linearity on the nonuniformity correction of IRFPA detection unit, the Kalman filter nonuniformity correction algorithm is extended and improved by using the non-linear model of detector response to effectively overcome the non-linear response Effect of correction accuracy. Firstly, the algorithm performs nonlinear compression on the image, converts it into a linear image, and then uses the Kalman filter algorithm in the linear model to perform non-uniform correction, and then takes an exponential operation to obtain a non-uniformly corrected original image. The experimental results show that this algorithm not only inherits the original algorithm to optimize the estimation of correction parameters by using scene information, but also solves the influence of detector offset and gain drift on calibration, and to some extent solves the problem of response nonlinearity correction Performance of the impact, resulting in better non-uniformity correction.