常用的红外热像仪信号处理方案主要有 3种，分立的 ARM或 DSP＋FPGA方案、TI达芬奇处理器方案和 FPGA方案。以上 3种方案分别存在芯片间带宽不足、不能支持焦平面探测器、串行处理不易实现等缺点。针对这些问题，提出一种基于 Xilinx ZYNQ处理器的红外热像仪信号处理方案，该方案选择非制冷红外焦平面探测器作为敏感元件。系统工作时，探测器产生的信号首先经过信号处理和模数转换操作，然后输入到 ZYNQ中，在 ZYNQ内部完成非均均匀性校正、温度测定和图像增强等处理，最后通过千兆以太网或者 HDMI接口输出。试验表明该方案数据带宽充足，能适配各种焦平面探测器或者红外机芯，串行处理易于实现。使用该方案设计的系统成像效果良好，测温精度满足要求。

Three signal processing schemes are widely used in infrared imaging systems. Their respective designs are based on separated ARM or DSP and FPGA, the TI Da-Vinci processor, and FPGA. All of these schemes have obvious defects, such as bandwidth shortage, incompatibility with infrared focal plane arrays (IRFPA), or inconvenience of implementing serial processing, etc. In this study, a signal processing schema based on the Xilinx ZYNQ processor is proposed to address these issues. The proposed schema leverages an uncooled focal plane array as the sensor. When the system initiates operation, the signal generated from the sensor is firstly transferred into ZYNQ after signal processing and analog-digital conversion. Subsequently, a series of operations including the non-uniformity correction (NUC), temperature measurement, and image enhancement are performed on the signal in ZYNQ. Finally, the corrected signal is output by the GigE interface or the HDMI port. A series of experiments demonstrate that the system designed from our proposed scheme is bandwidth-abundant, IRFPA-incompatible, and facile for the serial processing implementation. Our proposed system achieves attractive imaging performance and accurate temperature measurements.