F-LOAM
F-LOAM 以 LOAM 為基礎,著眼於降低計算量:運動畸變校正改為非迭代的兩階段方法,先以等速模型預測並校正,待位姿最佳化後再重算一次畸變並更新地圖。配準時把特徵點直接對齊全域邊緣地圖與平面地圖,並以局部平滑度作為權重,偏重跨幀較穩定的特徵。系統不含迴圈閉合,地圖以關鍵影格更新並做體素降採樣。
本頁內容
F-LOAM keeps LOAM features but replaces iterative distortion compensation with a two-stage constant-velocity scheme and weights point-to-edge/plane residuals by local smoothness when matching to global feature maps.
技術屬性
欄位內容為文獻擷取紀錄的原文用語(英文),以原文為據;「未查證」表示本研究尚未讀到該資訊,不代表該方法不具備此能力。
| 感測輸入 | 3D LiDAR only as method input: Velodyne HDL-64 as written (KITTI, whose cameras and GPS are not used), Velodyne VLP-16 on the warehouse AGV, virtual Velodyne VLP-16 in Gazebo; VICON motion capture used only as indoor ground truth |
|---|---|
| 原文測試平台 | vehicle (KITTI)、warehouse AGV with Intel NUC and Velodyne VLP-16 (model of AGV not stated)、simulation (virtual Pioneer robot in a Gazebo warehouse with moving workers, joystick-driven up to 2 m/s)、indoor VICON room |
| 狀態估計 | Gauss-Newton minimization of smoothness-weighted point-to-edge and point-to-plane distances to global feature maps (Sec. III-C) |
| 資料關聯 | LOAM-style smoothness features; global line/plane estimated from nearby map points by covariance eigen-analysis (KD-tree) (Sec. III-A, III-C) |
| 時間表示 | discrete scan poses; constant-velocity interpolation within a scan (Sec. III-B) |
| 去畸變 | non-iterative two-stage compensation: constant-velocity prediction before matching, then recomputation with the optimized pose before map update (Sec. III-B, III-D) |
| 迴圈閉合 | none |
| 全域最佳化 | none |
| 地圖表示 | global edge map and global planar map updated at keyframes and voxel-grid downsampled (Sec. III-D) |
| 先驗資訊 | none |
| 可輸出幾何 | global edge/planar feature maps and poses (Sec. III-D) |
| 計算需求 | more than 10 Hz on KITTI (Intel i7 3.2 GHz); Intel NUC on the AGV; ablation 69.07 ms/frame (Table I) |
使用設備
原文使用的感測器、運算硬體與載具(equipment)。型號保留原文寫法,連結到設備頁中同一型號的歸併名稱;角色依原文用途分為方法輸入、資料集感測器、執行運算平台、參考或真值量測(reference or ground truth)與比較對象設備。
| 類別 | 型號(原文寫法) | 角色 | 資料集 | 原文規格 | 出處 |
|---|---|---|---|---|---|
| LiDAR | Velodyne HDL-64歸入:Velodyne HDL-64E | 資料集感測器 | KITTI Odometry | 原文未報告 | (Wang et al., 2021a, Sec. IV-B) |
| LiDAR | Velodyne VLP-16 (virtual)歸入:Velodyne VLP-16 | 資料集感測器 | Gazebo simulated warehouse | Gazebo sensor model | (Wang et al., 2021a, Sec. IV-C1; Fig. 4b) |
| LiDAR | Velodyne VLP-16 | 方法輸入 | 未標示 | 原文未報告 | (Wang et al., 2021a, Sec. IV-C2) |
| GNSS 接收器 | KITTI GPS (model not stated) | 資料集感測器 | KITTI Odometry | not used by F-LOAM | (Wang et al., 2021a, Sec. IV-B) |
| 相機 | KITTI cameras (model not stated) | 資料集感測器 | KITTI Odometry | not used by F-LOAM | (Wang et al., 2021a, Sec. IV-B) |
| 載具平台 | Pioneer robot (virtual) | 資料集感測器 | Gazebo simulated warehouse | joystick-controlled, maximum 2 m/s | (Wang et al., 2021a, Sec. IV-C1; Fig. 4b) |
| 載具平台 | Automated Guided Vehicle for smart manufacturing (model not stated) | 方法輸入 | 未標示 | 原文未報告 | (Wang et al., 2021a, Sec. IV-C2; Fig. 5a) |
| 運算硬體 | Intel i7 3.2GHz processor based computer | 執行運算平台 | KITTI Odometry | used to time all compared methods on KITTI; ROS Melodic, Ubuntu 18.04 | (Wang et al., 2021a, Sec. IV-B) |
| 運算硬體 | Intel NUC mini computer | 執行運算平台 | 未標示 | onboard the AGV | (Wang et al., 2021a, Sec. IV-C2) |
| 其他 | VICON motion capture system歸入:Vicon motion capture system | 參考或真值量測 | 未標示 | indoor room ground truth | (Wang et al., 2021a, Sec. IV-C3) |
作者報告的優勢與限制
優勢
- On KITTI 00-10, with IMU and loop closure removed from all compared methods for consistency, F-LOAM achieved the second highest accuracy after IMLS-SLAM (whose results were copied from its paper) at more than 10 Hz, while LeGO-LOAM was the fastest
- the per-method values appear only in a bar chart
限制
- (verifier finding) The ablation (Table I) does not state the accuracy reference or the hardware, and the VICON test reports only an average error of 2 cm without RMSE, trajectory length or duration
營建工程相關證據
原論文測試於 KITTI、模擬倉儲與實體倉儲 AGV。Feng 等人(Feng et al., 2025)在施工中醫院門診大樓以預設參數測試,作者報告 F-LOAM 雖速度快,但在複雜施工場景精度下降,實際工地 APE RMSE 為 20.31 m(Table 4),並在三個與結構轉換相關的誤差峰中的第二、三個峰出現最高誤差(Sec. 5.3)。論文未說明實際工地 APE 所用參考軌跡的來源(全文僅描述 Gazebo 模擬的真實軌跡外掛),故實際工地 APE 只能視為作者報告值,不能當作已驗證的幾何精度。
原文驗證環境:公開基準、模擬、受控實驗、獨立參考量測
報告的性能數據
以下是原文作者報告的性能數值(author-reported results),不是本研究重新量測的結果。每張圖只並列同一個比較組(comparison group,同一張表、同一組實驗設定)內的方法;不同比較組之間的數值不可直接比較,也不構成排名。
本方法共出現在 39 個比較組,合計 220 筆紀錄。以下列出本方法紀錄最多的 4 組,其餘 35 組列在最後,並連到性能比較頁。
Ferrari et al., 2024 · Table II 本方法 19 筆
表格設定(擷取紀錄原文):KITTI benchmark RPE (%); segments 100-800 m for KITTI, MulRan and NC1, 10-80 m for NC0 and Hilti; averages exclude failures; per-sequence KITTI 00-10 and MulRan rows omitted (averages kept); values identical in arXiv v1 and version of record (Ferrari et al., 2024, Table II)
RPE [%] (segments 10-80 m),Newer College NC0 (OS0-128) · cat. easy
只並列這張表在相同設定下報告的方法;以「本方法:」開頭者為本頁方法。失敗、未執行與未報告以標記呈現,不是 0。
按 Tab 進入圖表後,用上下方向鍵逐一瀏覽各類別,Esc 關閉提示框;也可開啟表格檢視閱讀全部數值。
這些是 Ferrari et al., 2024 在此表設定下報告的數值(author-reported results),只能在同一個比較組內對照,不代表方法在其他資料或設定下的表現。
資料來源作者報告值(Ferrari et al., 2024, Table II)
| 方法(原文寫法) | 報告值 | 出處 |
|---|---|---|
| Ours (MAD-ICP)原文提出 | 1.16% | (Ferrari et al., 2024, Table II) |
| KISS-ICP | 2.06% | (Ferrari et al., 2024, Table II) |
| F-LOAM本方法 | 1.36% | (Ferrari et al., 2024, Table II) |
| MULLS | 2% | (Ferrari et al., 2024, Table II) |
| CT-ICP | 1.12% | (Ferrari et al., 2024, Table II) |
Vizzo et al., 2023 · Table III 本方法 16 筆
表格設定(擷取紀錄原文):MulRan; values are averages over the three runs per sequence; CT-ICP not evaluated because it lacks MulRan support (Vizzo et al., 2023, Table III)
Avg. tra. (KITTI relative translational error),MulRan · KAIST
只並列這張表在相同設定下報告的方法;以「本方法:」開頭者為本頁方法。失敗、未執行與未報告以標記呈現,不是 0。
按 Tab 進入圖表後,用上下方向鍵逐一瀏覽各類別,Esc 關閉提示框;也可開啟表格檢視閱讀全部數值。
這些是 Vizzo et al., 2023 在此表設定下報告的數值(author-reported results),只能在同一個比較組內對照,不代表方法在其他資料或設定下的表現。
資料來源作者報告值(Vizzo et al., 2023, Table III)
| 方法(原文寫法) | 報告值 | 出處 |
|---|---|---|
| MULLS [21] | 2.94% | (Vizzo et al., 2023, Table III) |
| SuMa [1] | 5.59% | (Vizzo et al., 2023, Table III) |
| F-LOAM [33]本方法 | 3.43% | (Vizzo et al., 2023, Table III) |
| Ours (KISS-ICP)原文提出 | 2.28% | (Vizzo et al., 2023, Table III) |
Palieri et al., 2021 · Table II 本方法 14 筆
表格設定(擷取紀錄原文):Husky field datasets from the SubT Urban (Alpha, Beta courses at the Satsop power plant) and Tunnel (Safety Research course, Bruceton mine) circuits; APE via evo against a reference from LOCUS scan matching on the DARPA ground-truth map; ME = RMSE of cloud-to-cloud error after ICP alignment of the map to the DARPA ground-truth map; loop closures disabled; FLOAM and LIO-Mapping ran with one LiDAR in Urban Alpha, LIO-SAM with one LiDAR; CPU loads from Urban Beta (LIO-SAM from Tunnel) (Palieri et al., 2021, Table II)
APE max,DARPA SubT Husky datasets (CoSTAR) · Urban Alpha course
只並列這張表在相同設定下報告的方法;以「本方法:」開頭者為本頁方法。失敗、未執行與未報告以標記呈現,不是 0。
- 失敗
按 Tab 進入圖表後,用上下方向鍵逐一瀏覽各類別,Esc 關閉提示框;也可開啟表格檢視閱讀全部數值。
這些是 Palieri et al., 2021 在此表設定下報告的數值(author-reported results),只能在同一個比較組內對照,不代表方法在其他資料或設定下的表現。
資料來源作者報告值(Palieri et al., 2021, Table II)
| 方法(原文寫法) | 報告值 | 出處 |
|---|---|---|
| LOCUS原文提出 | 1.69 m | (Palieri et al., 2021, Table II) |
| LOCUS FGA原文提出 | 0.63 m | (Palieri et al., 2021, Table II) |
| BLAM | 3.44 m | (Palieri et al., 2021, Table II) |
| ALOAM | 4.33 m | (Palieri et al., 2021, Table II) |
| FLOAM本方法 | 29.49 m | (Palieri et al., 2021, Table II) |
| Cartographer | 5.84 m | (Palieri et al., 2021, Table II) |
| LIO-Mapping | 2.12 m | (Palieri et al., 2021, Table II) |
| LIO-SAM | 無數值失敗註記(擷取紀錄):failed (authors could not get LIO-SAM working on the Urban datasets, likely because the 50 Hz IMU rate is below the recommended 200 Hz) | (Palieri et al., 2021, Table II) |
Li et al., 2021a · Table II 本方法 13 筆
表格設定(擷取紀錄原文):KITTI odometry 00-10; mean relative pose error over 100-800 m trajectories (rotation deg/100m / translation %); * marks sequences with loops; LOAM values quoted from its journal paper [19]; other baselines run with open-source code (Li et al., 2021a, Table II)
relative translational error (%),KITTI odometry · 00*
只並列這張表在相同設定下報告的方法;以「本方法:」開頭者為本頁方法。失敗、未執行與未報告以標記呈現,不是 0。
按 Tab 進入圖表後,用上下方向鍵逐一瀏覽各類別,Esc 關閉提示框;也可開啟表格檢視閱讀全部數值。
這些是 Li et al., 2021a 在此表設定下報告的數值(author-reported results),只能在同一個比較組內對照,不代表方法在其他資料或設定下的表現。
資料來源作者報告值(Li et al., 2021a, Table II)
| 方法(原文寫法) | 報告值 | 出處 |
|---|---|---|
| LOAM* (from [19]) | 0.78% | (Li et al., 2021a, Table II) |
| FLOAM本方法 | 0.92% | (Li et al., 2021a, Table II) |
| ISC-LOAM | 1.02% | (Li et al., 2021a, Table II) |
| SUMA | 0.77% | (Li et al., 2021a, Table II) |
| SUMA++ | 0.65% | (Li et al., 2021a, Table II) |
| Ours-ODOM原文提出 | 0.59% | (Li et al., 2021a, Table II) |
| Ours-LOOP原文提出 | 0.59% | (Li et al., 2021a, Table II) |
其他比較組
列出其餘 35 個比較組
- Chen et al., 2022b · Table II
- Zheng & Zhu, 2024 · Table I
- Zheng & Zhu, 2024 · Table III
- Feng et al., 2026 · Table 2
- Feng et al., 2026 · Table 6
- Pan et al., 2024 · Table VII
- Bavle et al., 2023 · Table II
- Pan et al., 2024 · Table VIII
- Zhu et al., 2025 · Table III
- Feng et al., 2025 · Table 3
- Feng et al., 2025 · Table 4
- Feng et al., 2026 · Table 1
- Feng et al., 2026 · Table 4
- Feng et al., 2026 · Table 5
- Bavle et al., 2023 · Table I
- Feng et al., 2026 · Table 8
- Lee et al., 2025a · Table II
- Pan et al., 2025 · Table III
- Wang et al., 2021a · Text Sec. IV-B
- Vizzo et al., 2023 · Table IV
- Palieri et al., 2021 · Table III
- Li et al., 2021a · Table IV
- Wang et al., 2021a · Table I
- Lee et al., 2025a · Table I
- Vizzo et al., 2023 · Table II
- Zheng & Zhu, 2024 · Table V
- Zhu et al., 2025 · Table I
- Zhu et al., 2025 · Table IV
- Wang et al., 2021a · Text Sec. IV-C3
- Lee et al., 2025a · Table III
- Ghadimzadeh Alamdari et al., 2025 · Table 3
- Ferrari et al., 2024 · Table III
- Zhang et al., 2024a · Table 8
- Zhang et al., 2024a · Table 9
- Zhu et al., 2025 · Table II
來源
Wang et al., 2021a
(2021)F-LOAM : Fast LiDAR Odometry and Mapping2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp. 4390-4396
DOI 10.1109/iros51168.2021.9636655arXiv 2107.00822程式碼
同儕審查已出版已讀全文近十年查證後修正
相關版本
- 預印本:arXiv 2107.00822 (title 'F-LOAM: Fast LiDAR Odometry And Mapping') https://arxiv.org/abs/2107.00822
- 程式碼釋出:wh200720041/floam https://github.com/wh200720041/floam
程式碼:https://github.com/wh200720041/floam(授權:BSD-style 3-clause text inherited from LOAM/A-LOAM (LICENSE file))。有公開程式碼不等於已被重現,也不代表目前版本與論文版本相同。