海天背景下舰船可见光图像目标区域定位方法

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。

Target Region Locating Algorithm for Ship Visual Image under Sea-sky Background

As for the row mean method has the disadvantages of low reliability for ship target location under sea-sky background,and easily influenced by the size of the ship target and complexity of sea-sky background,a new algorithm is proposed based on the line gray entropy for ship target region location. Based on the analysis of ship visual imaging characteristics under sea-sky background, the algorithm gives the concepts of image gray entropy and line gray entropy combined with information entropy theory. Using the line gray entropy to describe the degree of gray level change,variance of sliding interval is selected to search the sudden interval to determine the ship target region. Finally experimental verification is carried out by the actual image. Experimental results show that the algorithm can successfully locate the ship target region with higher accuracy and robustness than row mean method.