摘要
该系统在硬件上以大规模复杂可编程逻辑器件(FPGA)及微处理器(8051)为核心,采用模块化流水线处理结构,实现目标图像的实时跟踪。在系统软件上使用一种高效的快速相关算法MAD,建立相关跟踪置信度评估、模板自适应更换、目标丢失判断、抗干扰准则和再捕获等准则,有效地提高了目标跟踪的稳定性和抗干扰能力。
This system, with cores-large-scale Field programmed Gate Array and Single Chip Microprocessor (8051) as its core hardware, adopts the pipeline architecture to realize real-time image tracking. MAD, a fast image correlation tracking algorithm with excellent performance is used and a series of new criteria are constituted for tracking creditability evaluation, adaptive template change, similar object recognition and lost object acquisition, thus effectively improving the stability of object tracking and interference cancellation.
出处
《电子科技》
2004年第4期38-41,共4页
Electronic Science and Technology
