摘要
为准确实现船舶图像匹配,提出基于视觉传达技术的船舶图像匹配方法。考虑人眼的激励感知与抑制感知状态,在脉冲耦合神经网络内添设侧抑制信号等感知函数,实现低对比度的船舶图像增强处理;采用基于灰度差值直方图的特征提取方法判断增强后船舶图像边缘的粗细程度,提取船舶图像边缘特征;选取斯皮尔曼等级相关系数确定原始船舶图像与待匹配船舶图像的相关系数,以此判断船舶图像间的匹配度。测试结果表明,本文方法对简单与复杂背景的船舶图像均能实现图像增强,并充分获取图像内的边缘信息,实现准确的图像匹配。
In order to accurately realize ship image matching,a ship image matching method based on visual communication technology is proposed.Considering the excitation perception and suppression perception state of human eyes,the side suppression signal and other perception functions are added in the pulse coupled neural network to realize the low contrast ship image enhancement processing.The feature extraction method based on gray difference histogram is used to judge the thickness of the edge of the enhanced ship image and extract the edge features of the ship image.The Spearman rank correlation coefficient is selected to determine the correlation coefficient between the original ship image and the ship image to be matched,so as to judge the matching degree between ship images.The test results show that this method can enhance the ship images with simple and complex background,fully obtain the edge information in the image and realize accurate image matching.
作者
王丽莉
WANG Li-li(Jiangsu Maritime Institute,Nanjing 211170,China)
出处
《舰船科学技术》
北大核心
2022年第6期161-164,共4页
Ship Science and Technology
基金
中国交通教育研究会2020-2022年度交通教育科学研究课题(JTYB20-349)
江苏省现代教育技术研究课题(2021-R-91101)
江苏海事职业技术学院千帆团队建设资助项目。
关键词
视觉传达
船舶
图像增强
灰度差值
边缘特征提取
图像匹配
visual communication
warship
image enhancement
gray difference
edge feature extraction
image matching
