摘要
考虑到舰船检测问题为在平稳变化的场景提取稀疏小目标,引入了基于相位谱显著性的方法进行舰船目标检测.实验结果表明,此方法在计算速度上优于双参CFAR和Gabor滤波器方法,在检测性能上与Gabor滤波器方法相近,优于双参CFAR,能满足实际舰船检测系统的应用需求.
Considering ship detection as extracting sparse small targets in smooth background, a new ship target detection method based on phase spectrum is introduced. Experimental results show that the proposed algorithm gains better computing speed compared with dual parameters CFAR and Gabor filter algorithms, the performances of the proposed algorithm and Gabor filter algorithm are similar, both better than dual parameters CFAR algorithm. The proposed algorithm could satisfy the need of ship detection system.
出处
《沈阳大学学报(自然科学版)》
CAS
2015年第5期369-375,共7页
Journal of Shenyang University:Natural Science
基金
国家自然科学基金资助项目(61405041)
黑龙江省自然科学基金重点资助项目(ZD201216)
中国博士后科学基金资助项目(2014M551221)
关键词
舰船目标检测
光学遥感图像
显著性
相位谱
ship target detection
optical remote-sensing image
saliency
phase spectrum