为避免复杂海洋背景导致的舰船检测效果差问题，研究基于多特征融合的舰船目标检测方法。以选择性搜索算法获取初始舰船目标潜在区域为基础，结合几何和灰度特征约束从中选取舰船目标潜在区域，分别从舰船目标潜在区域中提取纹理、亮度、轮廓特征，通过自适应融合系数融合舰船目标多特征，以多特征融合结果为支持向量机分类器的输入，实现舰船目标检测。实验结果显示，该方法可有效降低38%以上的初始潜在区域数量；结合多特征的舰船目标描述能力，可实现精准舰船目标检测。
To avoid the problem of poor ship detection performance caused by complex ocean backgrounds, a ship target detection method based on multi feature fusion is studied. Based on the selective search algorithm to obtain the initial potential area of the ship target, combined with geometric and grayscale feature constraints, the potential area of the ship target is selected. Texture, brightness, and contour features are extracted from the potential area of the ship target, and the adaptive fusion coefficient is used to fuse the multiple features of the ship target. The fusion result of the multiple features is used as input for the support vector machine classifier to achieve ship target detection. The experimental results show that this method can effectively reduce the number of initial potential regions by more than 38%; By combining the ability to describe ship targets with multiple features, precise ship target detection can be achieved.
2024,46(9): 164-167 收稿日期：2023-09-19
DOI：10.3404/j.issn.1672-7649.2024.09.029
分类号：TP391.4
基金项目：教育部2019年第二批产学合作协同育人项目（201902005043）；四川省教育厅自然科学一般项目（18ZB0256）；四川省第二批地方普通本科高校应用型示范专业项目（255-256）
作者简介：刘玉洁（1985 – ），女，硕士，工程师，研究方向为图像识别及数据分析
参考文献：
[1] 杨龙, 苏娟, 黄华, 等. 一种基于深层次多尺度特征融合CNN的SAR图像舰船目标检测算法[J]. 光学学报, 2020, 40(2): 132-140.
YANG Long, SU Juan, HUANG Hua, et al. SAR ship detection based on convolutional neural network with deep multiscale feature fusion[J]. Acta Optica Sinica, 2020, 40(2): 132-140.
[2] 顾佼佼, 李炳臻, 刘克, 等. 基于改进Faster R-CNN的红外舰船目标检测算法[J]. 红外技术, 2021, 43(2): 170-178.
GU Jiao-jiao, LI Bing-zhen, LIU Ke, et al. Infrared ship target detection algorithm based on improved faster R-CNN[J]. Infrared Technology, 2021, 43(2): 170-178.
[3] 谢兆哲, 程永强, 吴昊, 等. 基于Toeplitz矩阵特征值分解的SAR图像舰船目标检测方法[J]. 信号处理, 2023, 39(3): 496-504.
XIE Zhao-zhe, CHENG Yong-qiang, WU Hao, et al. Ship target detection method in sar imagery based on eigenvalue decomposition of the toeplitz matrix[J]. Journal of Signal Processing, 2023, 39(3): 496-504.
[4] 吴天成, 王晓荃, 蔡艺军, 等. 基于特征融合的轻量级SSD目标检测方法[J]. 液晶与显示, 2021, 36(10): 1437-1444.
WU Tian-cheng, WANG Xiao-quan, CAI Yi-jun, et al. Lightweight SSD object detection method based on feature fusion[J]. Chinese Journal of Liquid Crystals and Displays, 2021, 36(10): 1437-1444.
[5] 严春满, 王铖. 基于选择性坐标注意力的SAR图像舰船目标检测[J]. 电子学报, 2023, 51(9): 2481-2491.
YAN Chun-man, WANG Cheng. Ship target detection in sar image based on selective coordinate attention[J]. Acta Electronica Sinica, 2023, 51(9): 2481-2491.
