摘要
海岸线与海天线检测作为前视红外成像型反舰导弹末制导技术中的关键技术,通常会受到岛岸、云层、亮带、条状波浪等多种因素干扰。为解决这一问题,提出了一种海岸线与海天线的通用检测方法。对原始图像构建积分图像,采用箱式滤波器来增强海岸线与海天线的边缘特征;逐行滑动统计矩形区域内像素的梯度显著性来确定海岸线与海天线潜在区域,通过潜在区域内逐列寻找显著性最大值点,并对所有的最大值点进行多项式迭代拟合,获得海岸线与海天线的准确位置;基于实际采集的前视红外海面场景图像对该方法进行了验证和分析。结果表明,通用检测方法能够克服岛岸、云层、亮带、条状波浪等复杂背景的干扰,实现海岸线与海天线的检测,场景适应性强,实时性好。
The coastline and sea-sky line detection is a key technology in the forward looking infrared( FLIR) terminal guidance technology of anti-ship missile,which is always interfered by island,cloud,bright band,strip wave and so on. A general detection method for coastline and sea-sky line is proposed.An integral image is constructed for the original image,and the box filter is used to enhance the edge features of coastline and sea-sky line. The gradient saliency of the pixels in the rectangular region is counted to determine the potential area of the coastline and sea-sky line. The points with maximum saliency are extracted in each column of the potential area and are used to obtain the locations of coastline and sea-sky line by polynomial iterative fitting. The proposed method is verified and analyzed based on the actual FLIR images of sea scene. The results show that the proposed method can overcome the interference of island,cloud,bright band,and strip wave to detect the coastline and sea-sky line. The proposed method has good scene adaptability and real-time performance.
作者
仇荣超
吕俊伟
宫剑
修炳楠
马新星
刘思彤
QIU Rongchao;LU Junwei;GONG Jian;XIU Bingnan;MA Xinxing;LIU Sitong(Naval Aviation University, Yantai 264001, Shandong, China;Special Service Department, Xi'an Flight Academy of Air Force, Xi'an 710300, Shaanxi, China)
出处
《兵工学报》
EI
CAS
CSCD
北大核心
2019年第6期1171-1178,共8页
Acta Armamentarii
基金
武器装备“十三五”预先研究项目(2016年)
关键词
前视红外图像
海岸线
海天线
梯度显著性
多项式迭代拟合
forward looking infrared image
coastline
sea-sky line
gradient saliency
polynomial iterative fitting
