基于卡尔曼与模糊PID的高帧频图像识别跟踪系统  

Image Recognition and Tracking System for High Frame Frequency Object Based on Kalman Predictor and Fuzzy PID Technology

作　　者：黄海波[1,2] 艾勇[1] 陈晶[1] 章勇勤[1] 左韬[1] 叶德茂[1] 

机构地区：[1]武汉大学电子信息学院,武汉430079 [2]湖北汽车工业学院,湖北十堰442002

出　　处：《光子学报》2009年第12期3295-3300,共6页Acta Photonica Sinica

基　　金：国家自然科学基金(10477014)资助

摘　　要：针对空间光通信系统对信标光斑实时准确跟踪需求,设计了基于高帧频相机的精跟踪处理控制平台,并将图像处理和控制算法全部集成在大规模现场可编程门阵列芯片内部.将模糊推理规则与常规比例、积分和微分算法结合,提高了非线性系统的控制准确度;将卡尔曼滤波器的预测功能与模糊比例、积分和微分结合起来,达到对快速倾斜镜的实时超前控制.现场实验表明,该系统能有效抑制外界干扰,实现对光斑弱小目标的准确跟踪,具有动态响应快、稳定准确度高和抗干扰能力强的特点.In order to tracking with the beacon facular real time and high precisely for fine tracking system in space light communication, a suit of fine tracking control flat based on high frame frequency CMOS camera is designed, image processing and control algorithms are all integrated into the large scale FPGA chip. Fuzzy reasoning rules are combined with general PID algorithm, which improve the control precision of nonlinear system; Kalman predictor is combined with fuzzy PID algorithm, which realizes real-time and forward control for FSM. Field experiment result shows that the system is capable of restraining outside disturbance, realizing exact tracking facular weak object, and it has the characteristics of high-speed response, high-precision stability and powerful ability of anti-jamming.

关 键 词：FPGA 卡尔曼预测 模糊PID 高帧频 

分 类 号：TN929.1[电子电信—通信与信息系统]