{
  "paper_id": "31796d480a9292f4310475cb2cecc938",
  "paper_title": "DUAL-Calib: Joint LiDAR-Dual-Camera Extrinsic Calibration with Automatic Target Corner Detection",
  "paper_authors": [
    "Jianyu Zhang",
    "Chong Wang",
    "Jiannan Chen"
  ],
  "publication_date": "",
  "publication_venue": "2026 5th Conference on Fully Actuated System Theory and Applications (FASTA)",
  "abstract": "This paper presents the DUAL-Calib calibration framework, which enables joint constrained extrinsic calibration of a single LiDAR and two cameras. As an optimized extension of FAST-Calib for multi-camera scenarios, DUAL-Calib establishes a unified optimization framework with inter-camera rigid geometric constraints to enhance calibration consistency and solve the cumulative error problem caused by independent calibration. Furthermore, this paper proposes an automatic corner detection algorithm based on ArUco marker triangulation, which avoids manual on-site measurement of the target's physical dimensions and enables automated field calibration. It can directly reconstruct the target coordinate system from the detected markers, realizing semi-automated calibration with reduced manual measurement. Experimental verification was performed on a wheel loader platform equipped with the Lingmou S1 perception module (integrating a Livox Mid360 LiDAR and two wide-angle fisheye cameras), with the calibration target placed within the overlapping field of view of the sensors at a distance of 1.5–4.5 m. The results demonstrate that the point-to-point registration error of DUAL-Calib on both camera channels is less than 5.0 mm under normal conditions and remains below 5.8 mm even under vibration interference. Compared with FAST-Calib's independent single-camera calibration method, DUAL-Calib reduces the calibration errors of LiDAR-Camera A and LiDAR-Camera B by 37.5% and 36.0% respectively with significantly lower variance, and the proposed automatic corner extraction method achieves sub-centimeter accuracy compared with manual measurement, effectively bridging the practicality gap between laboratory calibration and on-site deployment of construction machinery vehicles.",
  "keywords": [
    "LiDAR-Dual-Camera Calibration",
    "Extrinsic Calibration",
    "Construction Machinery",
    "Multi-Sensor Fusion"
  ],
  "paper_institutions": [
    "School of Electrical Engineering, Yanshan University, Qinhuangdao 066000, P. R. China"
  ],
  "output_language": "zh",
  "provider": "doubao",
  "model": "glm-5.3",
  "prompt_template": "simple",
  "extracted_at": "2026-09-17T09:41:25.030112Z",
  "template_tag": "simple",
  "summary": "### 摘要\n本文针对工程机械等现代机器人系统中单LiDAR与双相机外参标定存在的累积误差及人工测量依赖问题，提出DUAL-Calib联合约束标定框架。现有研究多聚焦单LiDAR-单相机标定（如FAST-Calib通过检测三维标靶圆孔实现毫米级精度），但独立标定多相机忽略了相机间刚性变换约束，且依赖预先人工测量标靶物理尺寸，难以满足施工现场快速重标定需求。DUAL-Calib在统一优化框架中引入以高精度预标定基线为核心的相机间刚性几何约束，并提出基于ArUco标记三角化的自动角点检测算法：通过PnP求解标记位姿、RANSAC拟合标定板平面、SVD闭式解重建标靶坐标系，从而免去现场人工测量。其核心机制为硬约束联合优化——仅优化 $T_{C_AL}$，由 $T_{C_BL}=T_{C_AC_B}T_{C_AL}$ 直接导出第二相机外参，并以Levenberg-Marquardt算法最小化全局点到点误差 $\\mathcal{E}=\\sum_{k\\in\\{A,B\\}}\\frac{1}{4N}\\sum_{j=1}^{N}\\sum_{i=1}^{4}\\|p_{i,j}^{C_k}-T_{C_kL}p_{i,j}^{L}\\|^2$。实验在装备Lingmou S1感知模块（Livox Mid360 LiDAR与两台1920×1200@10 Hz全局快门广角鱼眼相机）的轮式装载机平台上开展，标靶随机放置于传感器重叠视场内1.5–4.5 m处以模拟非标准化现场布局，并在光照变化、地面振动与背景杂乱条件下验证方法鲁棒性。结果表明：两相机通道点到点配准误差均低于5.0 mm，较FAST-Calib分别降低37.5%和36.0%且方差显著更小；振动干扰下误差仍低于5.8 mm；自动角点提取较人工测量达亚厘米精度（最大偏差0.28 cm）；重投影误差低于0.5像素；整体处理时间0.85 s、复杂度O(N)。未来工作将扩展至多相机系统、增强极端工程环境鲁棒性、集成IMU实现动态在线标定、面向边缘设备的轻量化优化以及多模态传感器联合标定。该方法适用于工程施工、农业机械与自动驾驶等需快速现场标定的场景，其前提约束为相机刚性安装于同一底盘且相机间基线经高精度预标定（平移误差<0.5 mm、旋转误差<0.1°）。\n\n### 关键要点\n- DUAL-Calib是FAST-Calib面向多相机场景的优化扩展，通过统一优化框架引入相机间刚性几何约束，解决独立标定导致的累积误差与结果不一致问题。\n- 提出基于ArUco标记三角化的自动角点检测算法，结合PnP位姿估计、RANSAC平面拟合与SVD闭式解直接重建标靶坐标系，免除人工现场测量标靶物理尺寸，实现半自动化标定。\n- 联合优化采用硬约束工程实现：仅优化 $T_{C_AL}$，第二相机外参由 $T_{C_BL}=T_{C_AC_B}T_{C_AL}$ 直接导出，以L-M算法最小化全局点到点配准误差，兼顾计算效率与物理一致性。\n- 相机间刚性基线 $T_{C_AC_B}$ 采用Zhang氏标定法离线预标定，平移误差优于0.5 mm、旋转误差优于0.1°，在刚性工程平台上被视为精确已知常量。\n- 跨相机一致性验证要求两相机重建的标记中心3D欧氏距离偏差小于0.5 mm阈值，不满足时需复查或人工干预，以确保自动提取特征在几何配准前的可靠性。\n- 实验在装备Lingmou S1模块（Livox Mid360 LiDAR+两台全局快门鱼眼相机）的轮式装载机上开展，标靶随机放置于1.5–4.5 m重叠视场内，模拟非标准化现场布局并验证光照变化、振动与背景杂乱下的性能。\n- 与FAST-Calib各10组数据对比：DUAL-Calib的LiDAR-相机A/B平均误差为4.5±0.7 mm与4.8±0.8 mm，分别降低37.5%和36.0%且方差显著更低，总体处理时间0.85 s（FAST-Calib为0.68 s）。\n- 自动角点检测与人工标注对比：左/右相机最大偏差0.26/0.28 cm，平均偏差0.11/0.13 cm，标准差约0.03–0.04 cm，达到亚厘米级精度。\n- 重投影验证显示左/右相机平均重投影误差为0.32/0.35像素，最大值控制在0.5像素内，波动范围小于0.15像素，验证了标定结果的高精度与稳定性。\n- 振动条件下（怠速及≤5 km/h低速行驶）DUAL-Calib平均配准误差保持在5.8 mm内，平移偏差<0.5 mm、旋转偏差<0.1°，而FAST-Calib独立模式最大偏差达2.3 mm和0.5°。\n- 方法整体时间复杂度为O(N)，独立注册经两次Kabsch-SVD求解初值以防联合优化陷入局部最优，总运行时间低于1秒，满足现场快速标定需求。\n- 适用约束：需相机刚性安装于同一底盘以保持固定相对位姿，并依赖高精度预标定基线；未来将扩展多相机系统、集成IMU动态在线标定并面向边缘设备轻量化部署。",
  "available_templates": [
    "deep_read",
    "simple"
  ]
}