摘要
针对行均值法用于海天背景下舰船目标区域定位可靠性不高、容易受到舰船目标大小和海天背景复杂程度影响的不足,提出一种基于图像行灰度熵的舰船目标区域定位方法。该方法在分析海天背景下舰船目标可见光图像成像特点基础上,结合信息熵理论给出了图像灰度熵及行灰度熵的概念,利用图像行灰度熵描述图像在行方向上的灰度变化程度,选用滑动区间方差搜索策略得到图像行灰度熵曲线的突变区间,从而实现舰船目标区域定位,最后通过实际图像进行了实验验证。实验结果表明,该方法能对舰船目标所在区域有效定位,其准确性及鲁棒性均优于行均值法。
As for the row mean method has the disadvantages of of low the reliability ship for and for ship target location of under sea-sky background,and easily influenced proposed by the on size the target entropy complexity ship sea-sky background,a new algorithm is based line gray target region the location.Based gives on the analysis of image ship visual imaging characteristics line under sea-sky background,algorithm the concepts the of gray entropy and gray entropy of combined level with information entropy sliding theory.interval Using is line gray entropy to describe the degree gray the change,variance target of selected to search carried the out sudden by interval actual to determine ship region.show Finally that experimental algorithm verification is the image.with Experimental results and the row can successfully locate the ship target region higher accuracy robustness than mean method.
出处
《火力与指挥控制》
CSCD
北大核心
2017年第7期66-71,共6页
Fire Control & Command Control
基金
海军预研基金资助项目(435515908)
关键词
舰船定位
行灰度熵
海天背景
可见光图像
ship location
line gray entropy
sea-sky background
visual image
