刘博文¹, 戴永寿¹, 金久才², 李立刚¹, 邵 峰^1,2
1.中国石油大学(华东);2.国家海洋局 第一海洋研究所

摘要:
海面运动载体获取的可见光图像中, 远景目标成像面积小、特征不显著、信噪比较低等特点增大了目标检测难度, 现有方法难以同时取得较高的查全率与较低的虚警率。作者根据目标在图像中的空间分布与统计特性, 提出了一种海面远景目标检测方法。首先结合直线特征与界线特征实现海天线检测, 然后根据拉伊达准则实现明度图像中疑似目标像素与背景像素的分离, 得到二值图像, 最后对二值图像进行形态学处理, 并根据疑似目标区域与海天线的位置关系, 排除干扰区域。测试结果表明:所提方法具有较高查全率与较低虚警率, 且实时性和鲁棒性较强。
关键词:海面远景目标检测海天线检测拉伊达准则形态学处理
DOI：10.11759/hykx20171011005
分类号:
基金项目:国家重点研发计划项目(2017YFC1405203); 国家自然科学基金资助项目 (61401111)

Marine farsighted target-detection method based on spatial distribution and statistical characteristics
LIU Bo-wen,DAI Yong-shou,JIN Jiu-cai,LI Li-gang,SHAO Feng
Abstract:
In visible-light maritime images, the area and characteristics of farsighted targets are not clear and the image SNRs are low. The existing target-detection method cannot obtain a high recall and low false alarm rate. Based on the target’s spatial distribution and statistical characteristics in the image, we propose a marine farsighted target-detection method. In this method, first, the sea–sky line is detected by combining the line and boundary features. Then, the Pauta criterion is used to obtain a binary image by separating the suspected target pixel from the background pixel. Finally, the binary image is morphologically processed and the interference area excluded based on the positional relationship between the suspected target area and the sea–sky line. Our experimental results indicate that the proposed method has a high recall rate and low false alarm rate, and is relatively robust and effective in real time.
Key words:marine farsighted target detectionsea–sky line detectionPauta criterionmorphological processing





PDF全文下载地址:

点我下载PDF