摘要
为了利用光学遥感影像进行舰船目标型号的识别,介绍了一种基于SIFT(Scale Invariant Feature Transform)特征和Harris角点特征匹配的舰船型号识别方法.首先,利用SIFT算子提取舰船局部特征;为减少SIFT特征点误匹配,对舰船进行分区域(舰首、舰舯、舰尾)匹配.其次,提取描述舰船边缘的Harris角点特征,进行匹配.最后,根据2次匹配结果的重要性分别赋予相应的权重,进而判定舰船的型号.针对护卫舰的实验表明,该方法能够有效地对光学遥感影像的舰船型号进行识别.
To identify the type of a ship in optical remote sensing images, a ship recognition method based on feature matching was proposed, including SIFT （scale invariant feature transform） and Harris corner. First, the local features of a ship were extracted by the SIFT operator. To reduce the mismatching of SIFT, a ship was divided into three regions （the bow, the ship amidships and the stern） and then the cor- responding region was matched respectively. Next, the features of the Harris corner was extracted, which described the edge of the ship, and then the matching was performed. Finally, different weights according to the importance of the two matches were given, and the type of a given ship was determined. The experimental results on frigates shows that the method can effectively identify the type of a given ship in optical remote sensing images.
出处
《上海交通大学学报》
EI
CAS
CSCD
北大核心
2016年第9期1415-1421,共7页
Journal of Shanghai Jiaotong University
基金
国家自然科学基金资助项目(U1406404
41174164
61331015)
关键词
舰船型号识别
光学遥感影像
SIFT特征
HARRIS角点
图像匹配
ship recognition
optical remote sensing image
scale invariant feature transform （SIFT） fea- tures
Harris corner
image matching
