弹道目标识别的红外辐射数据仿真研究

充足的目标红外辐射数据集是研究弹道目标检测、跟踪与识别的必备前提，以往弹道目标红外数据仿真模型强调特定参数场景下的光学成像研究，以解决目标检测和关联跟踪等图像信息处理问题；而目标识别所需的数据仿真模型更注重在可能的动态参数范围内对目标辐射强度信号变化规律的探究。为了满足目标识别的数据仿真要求，着重探讨了目标场景参数和辐射强度信号动态变化特性；并根据仿真结果分析了影响数据差异特性的底层因素，为设计适合任意仿真场景数据的目标识别算法予以启示。

Research on infrared irradiance data stimulation for ballistic targets recognition

Adequate IR data set are essential to design the algorithms of ballistic targets detection, trace and recognition. In traditional IR image stimulations, the systems often emphasize optical imaging research of some specific scenario in order to solve the related image processing problems involving targets detection and data association. But for target recognition, the required data stimulation model would place emphasis on the study of change rule of radiant power in all of the potential dynamic parameters. To satisfy the requirements of data simulation for target recognition, the research emphasis in this paper was put on the study of scenario parameters and radiant dynamic characteristics. Meanwhile, the underlying factors were analyzed for the aid to design the widely-used recognition algorithm.