基于FPGA的高速实时图像采集和自适应阈值算法  

High-speed real-time image acquisition and auto-adapted threshold processing based on FPGA

作　　者：李冬梅[1] 王延杰[2] 王长宇[1] 陈枫[1] 李德辉[3] 

机构地区：[1]空军航空大学信息对抗系,长春130022 [2]中国科学院长春光学精密机械与物理研究所图像室,长春130033 [3]中国电子科技集团公司第二十六研究所,重庆400060

出　　处：《吉林大学学报（工学版）》2010年第2期534-538,共5页Journal of Jilin University:Engineering and Technology Edition

基　　金：'863'国家高技术研究发展计划项目(2006AA703405F)

摘　　要：提出了基于FPGA的图像处理自适应阈值算法,实现了激光光斑中心的高速实时检测。采用3×3窗口模块和自适应阈值模块,先对CCD输入数据进行处理,判断光斑的范围,然后再运用光斑的质心算法对光斑所占的像元进行运算,得出光斑位置的脱靶量。本文达到了脱靶量帧速3000帧/s、精度2μrad的技术指标,实现了高速率、高精度的跟踪要求。An auto-adapted threshold algorithm for image processing is proposed,which is based on Field Programmable Gate Array（FPGA）.Using this algorithm can realize high-speed real-time detection for laser spot center.3×3 window module and auto-adapted threshold module are adopted in this study.First the CCD input data are processed,and the spot scope is estimated.Then the spot centroid algorithm is used to process the pixels,which are occupied by the spot,and the miss distance of the spot position is obtained.Finally,the image is displayed on the LCD using VGA format.This algorithm can reach miss distance frame rate of 3000 F/s and accuracy of 2 μrad.By this proposed method we can realize high-speed and high accurate tracking.

关 键 词：信息处理技术 FPGA CAMERA Link接口协议 自适应阈值 图像采集 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]