论文部分内容阅读
德国AEG红外组件GmbH(AIM)公司最近研制成的几种新型二维探测组件,已使其焦平面列阵家族得到了扩大。该公司已推出了大型量子阱和碲镉汞器件,例如,一种 512×640量子阱长波红外组件。这些器件都能提供高的热分辨率,其典型噪声等效温差(NETD)低于25 mK。在这个性能范围内,非均匀性校正(NUC)后的剩余空间噪声往往就是极限噪声源。为了在时间噪声受限制的条件下操作,必须采用比二点校正更高级的校正算法。AIM公司已开发成了一种符合这些要求的自适应迭代校正算法(SAICA)。其基于场景的逼近免除了采用热参考的定标步骤。这种算法是在AIM公司的标准视频图像处理机VIP-ICM80上实施的,运行时所用的最大帧速为50 Hz,图像尺寸为 256像元×256像元。AIM公司的这些新型探测组件的较大图像尺寸和较高帧速,促使其设计出了一种新的视频图像处理板。该处理板还提供了一个微型扫描仪接口,它可以用来提高凝视红外成像组件的几何分辨率。 本文介绍了AIM公司二维探测组件用的非均匀性校正硬件和软件的实施情况以及测得的有关这些器件的校正能力的性能数据。
Several new two-dimensional probe assemblies recently developed by AEG Infrared Components GmbH (AIM) of Germany have expanded their family of focal plane arrays. The company has introduced large quantum wells and mercury cadmium telluride devices, for example, a 512 × 640 quantum well long-wave infrared module. These devices provide high thermal resolution with a typical NETD of less than 25 mK. Within this performance range, residual noise after non-uniformity correction (NUC) is often the source of extreme noise. In order to operate under time-limited noise conditions, a more advanced correction algorithm than two-point correction must be used. AIM has developed an adaptive iterative correction algorithm (SAICA) that meets these requirements. Its scene-based approximation eliminates the scaling step with thermal reference. The algorithm is implemented on AIM’s standard video processor VIP-ICM80 with a maximum frame rate of 50 Hz and an image size of 256 pixels × 256 pixels. The larger image size and higher frame rates of these new detector assemblies at AIM led to the design of a new video processing board. The processing board also provides a micro-scanner interface that can be used to increase the geometric resolution of the staring infrared imaging assembly. This article describes the implementation of non-uniformity correction hardware and software used by AIM’s 2D detection components and measured performance data on the calibration capabilities of these devices.