CAJ | 학술논문

针对导弹末制导红外图像处理实时性要求,分析了红外图像并行实时处理算法及其集群计算机的实现,建立了适应于实时处理的BSP改进模型和集群并行处理软件系统,并给出了红外图像并行处理算法的实现方法.研究了应用提升小波实现图像并行处理算法及实现,进而分析了算法性能,推导出了加速比模型.实验验证了其有效性.
침대도탄말제도홍외도상처리실시성요구,분석료홍외도상병행실시처리산법급기집군계산궤적실현,건립료괄응우실시처리적BSP개진모형화집군병행처리연건계통,병급출료홍외도상병행처리산법적실현방법.연구료응용제승소파실현도상병행처리산법급실현,진이분석료산법성능,추도출료가속비모형.실험험증료기유효성.
Considering the real-time requirement of processing infrared images on missile flight in end-game,this paper makes an analysis on parallel algorithms for the real-time processing of infrared images and on the realization of cluster computers.The BSP algorithm model for real-time processing and the cluster software system for parallel processing have been set up,with the suggested means of realization of parallel algorithms for processing infrared images.Meanwhile,it also makes a research on the parallel algorithms by application of Lifting Wavelet Transform as well as.their realization,analyzing hence the performance of the Lifting Scheme with derivation of the speedup-ratio model.In addition,the model's validity has been verified by experiments.