毫米波雷达、激光雷达能够给出相对准确的目标距离、尺寸信息，而视觉系统能够给出较直接且丰富的目标纹理特征，将雷达与视觉结合可有效提升无人船障碍物检测性能。本文提出一种基于目标距离及尺寸因子的多传感器数据融合方法。首先，采用串联式目标匹配方法，完成双雷达以及与单目相机的障碍物匹配。其次，利用目标距离及尺寸因子计算各传感器的检测信任度，然后利用Dempster-Shafer(D-S)证据理论对检测信任度进行融合，完成虚假目标剔除。最后，通过无人船海上近距离、远距离和多目标场景试验验证了本文方法的有效性。
Millimeter-wave radar and lidar can give relatively accurate target distance and size information, while vision systems can give more direct and rich target texture features. Combining radar and vision can effectively improve obstacle detection performance of unmanned surface vessel(USV). This paper presents a multi-sensor data fusion method based on target distance and size factors. Firstly, the detection data of the same obstacle in the detection results of the three sensors are obtained by using the tandem target matching method. Secondly, the detection confidence of each sensor is calculated by the target distance and size factors. Then, the Dempster-Shafer (D-S) evidence theory is used to fuse the detection confidence of each sensor, and the target lower than the fusion confidence threshold is treated as a false target. Finally, the effectiveness of the proposed method is verified by measured data. The results show that the proposed method can effectively reduce the false alarm rate and missed detection rate of single sensor.
2023,45(20): 87-92 收稿日期：2022-7-11
DOI：10.3404/j.issn.1672-7649.2023.20.016
分类号：U675.79
基金项目：国家重点研发计划项目（2021YFC3101101）
作者简介：葛燕龙(1997-),男,硕士研究生,研究方向为无人船多传感器数据融合
参考文献：
[1] 朱健楠, 虞梦苓, 杨益新. 无人水面艇感知技术发展综述[J]. 哈尔滨工程大学学报, 2020, 41(10): 1486–1492
[2] HERMANN D, GALEAZZI R, ANDERSEN J C, et al. Smart sensor based obstacle detection for high-speed unmanned surface vehicle[J]. IFAC-Papers Online, 2015, 48(16): 190–197
[3] CHENG Y W, XU H and LIU Y M. Robust small object detection on the water surface through fusion of camera and millimeter wave radar[C]//2021 IEEE/CVF International Conference on Computer Vision (ICCV). Montreal, QC, Canada: IEEE, 2021: 15243–15252.
[4] ZHANG X L, WANG H S, CHENG W D. Vessel detection and classification fusing radar and vision data[C]//7th International Conference on Information Science and Technology (ICIST), Da Nang, VIETNAM: IEEE, 2017: 474–479.
[5] LIU H, NIE J, LIU Y J, et al. A multi-modality sensor system for unmanned surface vehicle[J]. Neural Processing Letters, 2020, 52(2): 977–992
[6] LIU W W, LIU Y C, GUNAWAN B A, et al. Practical moving target detection in maritime environments using fuzzy multi-sensor data fusion[J]. International Journal of Fuzzy Systems, 2020: 1–19.
[7] CHEN J Y, WANG H A. Obstacle detection method for usv by fusing of radar and motion stereo[C]//2020 IEEE 16th International Conference on Control & Automation (ICCA). Singapore: IEEE, 2020: 159–164.
[8] 柳毅, 佟明安. 匈牙利算法在多目标分配中的应用[J]. 火力与指挥控制, 2002, 27(4): 34–37.
[9] 张翠文, 张长伦, 何强, 等. 目标检测中框回归损失函数的研究[J]. 计算机工程与应用, 2021, 57(20): 97-103.
[10] 韩德强, 杨艺, 韩崇昭. DS证据理论研究进展及相关问题探讨[J]. 控制与决策, 2014, 29(01): 1-11.
[11] 李瑞伟, 李立刚, 金久才, 等. 基于欧氏距离的无人艇载毫米波雷达点迹凝聚方法[J]. 水下无人系统学报, 2020, 28(6): 604-610.
