Abstract: In order to meet the quasi-real time demands in the application of military remote sensing imagery on the domestic independent and controllable software and hardware platform, it analyses the difference between independent and controllable computer hardware and software and that of commercial system. It researches the hardware and software cross platform compatibility technology. Remote sensing imagery auto-processing system architecture with high cohesion and low coupling is designed based on layered design. Cross platform consistent source code technology is proposed. Large scale remote sensing imagery auto management and high performance computing based on open source Hadoop are realized. Cross platform remote sensing imagery processing system based on independent and controllable computer hardware and software is constructed. The remote sensing images can be processed real-time on the system. The ability of cross platform application software development is significantly improved by the proposed method. The stability of the system is improved too. The method can be applied to the development of other similar systems.

Key words: independent and controllable platform, cross platform, layered design, source code consistent development, remote sensing image data automatically processing

摘要： 针对军事遥感影像应用时效性的需求以及在自主可控国产软硬件环境下开发军事应用系统的需要，对国产软硬件环境与主流商用系统差异进行了分析，在此基础上对软、硬件兼容技术进行了研究，基于分层思想设计了高内聚、低耦合的遥感影像处理系统，提出源代码一致的跨平台程序设计方法，基于开源Hadoop平台实现了遥感数据管理与高性能计算，并在此基础上生成基于国产软硬件环境下的遥感影像自动处理系统，实现了遥感影像准实时处理，运用于军事应用系统。该跨平台系统设计与实现方法显著提高跨平台应用软件开发效率，且有利于提高系统的稳定性，可以应用于其他类似跨平台系统设计开发。

关键词: 自主可控平台, 跨平台, 分层设计, 源代码一致, 遥感影像自动处理