基于区域外接矩形的自动化孔洞填充算法

Automatic Hole-Filling Algorithm Based on Region External Rectangle

基于图像进行船舶目标检测与识别时,船舶目标区域内部常出现孔洞,从而降低船舶目标检测与识别的准确率和稳定性。针对该问题,提出了一种用于二值图像的基于船舶目标区域外接矩形的内部孔洞自动填充算法。首先,寻找目标区域最小外接矩形,并进行标注;然后,引入边界限定算法对内部孔洞自动填充。该算法抗干扰能力强,不受船舶区域内部孔洞大小及数目的限制;同时,对于由船舶灰度不均导致图像分割时产生的船舶残缺具有修复作用。试验结果表明,该算法可以有效填充船舶目标区域内部孔洞,为准确提取船舶目标特征以及提高船舶目标检测与识别的准确率和稳定性提供支撑。

During boat target detection and recognition based on images, holes often appear in the boat target region. Thus, the accuracy and the stability of boat target detection and recognition are dropped. Aimed at the problem, an automatic interior hole-filling algorithm based on the ex-ternal rectangle of the boat target regions is proposed for binary images. Firstly, the minimum external rectangle of the target region is searched and marked. Then, the boundary limitation al-gorithm is presented to automatically fill the interior holes. The algorithm has strong anti-inter-ference ability and cannot be restricted by the size and the amount of the interior holes of the boat target region. Meanwhile, it has repair function to the situation of deformed boat caused by the uneven grey scale of the boat during the image division. Experimental result shows that the algo-rithm can effectively fill the interior holes of the boat target region, thus providing support for accurately extracting boat target features and enhancing the accuracy and the stability of boat tar-get detection and recognition.