摘要
描述了在自主研制的TMS320DM642图像融合系统平台上,使用实时多线程操作系统(DSP/B IOS)开发驱动程序和编写融合算法的方法和流程。综合运用了B IOS、CSL、BSL、DDK、RF5、XD IAS等技术编写设备驱动程序,使用标准的类/微型驱动程序模型开发新的视频驱动程序,实现了软件开发的模块化和层次化。在此基础上编写了图像融合程序,总结了CCS编程中使用的各项优化技术。与传统软件开发比较,基于DSP/B IOS的DSP驱动程序缩短了开发时间,增加了代码的兼容性、继承性和可移植性。
In the paper the method and flow of driver development are described using the real-time multi-threading operating system DSP/BIOS and image fusion algorithms on the independent development image fusion system based on TMS320DM642. The TI techniques, such as BIOS, chip support library （ CSL）, board support library（ BSL）, driver development kit （DDK）, reference frameworks level 5 （RF5）, expressDSP algorithm standard （XDIAS）, etc, are employed synthetically to develop the device driver. The standard class/mini-driver model is used to develop the new video driver. It makes the modularization and hiberarchy of development realizable. On the bais the fusion algorithms are programmed and the optimization techniques in CCS programming are summarized. Compared with traditional software development, the driver development based on DSP/BIOS shortens development time and increases compatibility, inherited attribute and portability.
出处
《激光与红外》
CAS
CSCD
北大核心
2006年第11期1070-1074,共5页
Laser & Infrared
基金
973项目资助
