摘要
针对海洋背景下海面舰船检测存在的问题,本文提出了一种海洋背景下海面舰船快速检测方法,该方法通过计算图像的高斯金字塔高频分量,得到三层拉普拉斯降维图,第二层降维图与原图相比,维数降低了16倍;对降维图通过改进Sobel检测,增强降维图中舰船目标对比度,突出感兴趣区域,实现快速检测。实验结果表明,该方法具有一定的鲁棒性,能够实现海面舰船的快速检测。
Aiming at the detection of the ship on sea surface under the ocean background,a fast detecting algorithm for ships on sea surface is proposed.The high frequency component of Gaussian pyramid in the image is calculated,and three layer dimension reducing images of Laplacian pyramid are obtained.Compared with the original image,the dimensionality of the second layer dimension reducing image is reduced by 16 times.After that,based on the improved Sobel edge detection in dimension reducing image,the contrast of ships is enhanced and the region of interest is highlighted to achieve rapid detection.The experiment results show that this method has certain robustness and can achieve the fast detection of ships on sea surface.
出处
《激光与红外》
CAS
CSCD
北大核心
2016年第5期602-606,共5页
Laser & Infrared
基金
国家自然科学基金项目(No.61171170)
安徽省自然科学基金项目(No.1408085QF115)资助
关键词
SOBEL算法
高斯金字塔
降维图
感兴趣区域
Sobel algorithm
Gaussian pyramid
reduce dimension image
region of interest（ROI）
