摘要
为提高动态环境下对“猫眼”目标进行激光主动探测的时效性和准确性,提出并验证了基于OFSD配准算法的“猫眼”目标识别方法。识别“猫眼”目标的过程中:对探测到的主被动图像进行预处理,降低数据量和降低光照不均匀的影响,并通过OFSD方法进行图像配准,配准后的两帧图像进行帧间互差差分运算及二值化处理,对处理后的图像提取形态和灰度特征进行“猫眼”目标识别。实验结果表明,基于OFSD配准算法的“猫眼”目标识别方法在动态环境中能够有效快速地进行“猫眼”目标识别,大大提高了“猫眼”目标识别算法的实时性。
In order to improve the timeliness and accuracy of laser active detection of“cat's eye”target in dynamic environment,the“cat's eye”target recognition method based on OFSD registration algorithm was proposed and verified.In the process of identifying the“cat's eye”target,the detected active and passive images were pre-processed to reduce the amount of data and reduce the influence of uneven illumination;and the image registration was performed based on OFSD method,differential processing and binarization processing were carried out with the two registered images;and the“cat's eye”target recognition was performed according to the morphology and grayscale features extracted from the processed images.The experimental results show that the“cat's eye”target recognition method based on OFSD registration algorithm can effectively and quickly perform“cat's eye”target recognition in dynamic environment,which greatly improves the real-time performance of“cat's eye”target recognition algorithm.
作者
王喆堃
朱精果
姜成昊
解天鹏
WANG Zhe-kun;ZHU Jing-guo;JIANG Cheng-hao;XIE Tian-peng(Academy of Opto-Electronics,Chinese Academy of Sciences,Beijing 100094,China;Institute of Microelectronics,Chinese Academy of Sciences,Beijing 100029,China;University of Chinese Academy of Sciences,Beijing 100039,China)
出处
《计算机仿真》
北大核心
2020年第8期414-418,共5页
Computer Simulation
基金
国家自然科学基金资助项目(61605216)。
关键词
“猫眼”目标
激光主动探测
图像配准
目标识别
“Cat's eye”target
Laser active detection
Image registration
Target identification