摘要
传统船舶特征提取算法,无法完全对船舶特征进行提取,存在精度不够、无法有效提取特征点等问题。为有效解决此问题,提出基于图像检测的船舶特征提取优化算法。计算不同波段下船舶图像的灰度值,依据灰度值结合霍夫曼及分裂排序编码构建船舶图像特征提取示意图;设计基于图像的船舶特征点检测提取模型,确定图像中船舶特征点相关性。依据相关性进行船舶特征点的有效提取,并利用图像特征提取,完成船舶特征提取的优化算法,设计对比实验结果表明,改进后方法与传统方法相比,大幅提高船舶特征提取的准确性。
The traditional ship feature extraction algorithm can not extract the ship features completely, and there is not enough precision to extract feature points effectively. In order to solve this problem effectively, a ship feature extraction optimization algorithm based on image detection is proposed. The data format model of ship feature points based on image is constructed to determine the correlation of ship feature points in images. Basis of correlation ship of feature points extracted effectively, and make use of image feature extraction, complete ship optimization algorithm of feature extraction, design compared the experimental results show that the improved method is compared with traditional methods, significantly improve the accuracy of the ship's feature extraction.
出处
《舰船科学技术》
北大核心
2018年第5X期175-177,共3页
Ship Science and Technology
关键词
船舶特征
图像检测
方法改进
数据提取
相关性
characteristics of ship
image detection
method improvement
data extraction
correlation
